Pass formattable object to logger instead of intermediate string

Instead of using std::function<std::string()> which creates intermediate
strings, introduce allocating_section_namer class that formats directly
to fmt::memory_buffer. This avoids string allocation overhead when
logging allocation failures.

Co-authored-by: tgrabiec <283695+tgrabiec@users.noreply.github.com>

.github/CODEOWNERS

@@ -1,5 +1,5 @@
 # AUTH
-auth/* @nuivall
+auth/* @nuivall @ptrsmrn
 
 # CACHE
 row_cache* @tgrabiec
@@ -25,11 +25,11 @@ compaction/* @raphaelsc
 transport/*
 
 # CQL QUERY LANGUAGE
-cql3/* @tgrabiec @nuivall
+cql3/* @tgrabiec @nuivall @ptrsmrn
 
 # COUNTERS
-counters* @nuivall
-tests/counter_test* @nuivall
+counters* @nuivall @ptrsmrn
+tests/counter_test* @nuivall @ptrsmrn
 
 # DOCS
 docs/* @annastuchlik @tzach
@@ -57,6 +57,7 @@ repair/* @tgrabiec @asias
 
 # SCHEMA MANAGEMENT
 db/schema_tables* @tgrabiec
+db/legacy_schema_migrator* @tgrabiec
 service/migration* @tgrabiec
 schema* @tgrabiec
 

.github/scripts/auto-backport.py

@@ -62,7 +62,7 @@ def create_pull_request(repo, new_branch_name, base_branch_name, pr, backport_pr
         if is_draft:
             labels_to_add.append("conflicts")
             pr_comment = f"@{pr.user.login} - This PR was marked as draft because it has conflicts\n"
-            pr_comment += "Please resolve them and remove the 'conflicts' label. The PR will be made ready for review automatically."
+            pr_comment += "Please resolve them and mark this PR as ready for review"
             backport_pr.create_issue_comment(pr_comment)
         
         # Apply all labels at once if we have any

.github/workflows/backport-pr-fixes-validation.yaml

@@ -18,7 +18,7 @@ jobs:
             
             // Regular expression pattern to check for "Fixes" prefix
             // Adjusted to dynamically insert the repository full name
-            const pattern = `Fixes:? ((?:#|${repo.replace('/', '\\/')}#|https://github\\.com/${repo.replace('/', '\\/')}/issues/)(\\d+)|([A-Z]+-\\d+))`;
+            const pattern = `Fixes:? (?:#|${repo.replace('/', '\\/')}#|https://github\\.com/${repo.replace('/', '\\/')}/issues/)(\\d+)`;
             const regex = new RegExp(pattern);
             
             if (!regex.test(body)) {

.github/workflows/call_sync_milestone_to_jira.yml

@@ -1,14 +0,0 @@
-name: Call Jira release creation for new milestone
-
-on:
-  milestone:
-    types: [created]
-
-jobs:
-  sync-milestone-to-jira:
-    uses: scylladb/github-automation/.github/workflows/main_sync_milestone_to_jira_release.yml@main
-    with:
-      # Comma-separated list of Jira project keys
-      jira_project_keys: "SCYLLADB,CUSTOMER"
-    secrets:
-      caller_jira_auth: ${{ secrets.USER_AND_KEY_FOR_JIRA_AUTOMATION }}

.github/workflows/docs-validate-metrics.yml

@@ -7,7 +7,7 @@ on:
       - enterprise
     paths:
       - '**/*.cc'
-      - 'scripts/metrics-config.yml'
+      - 'scripts/metrics-config.yml' 
       - 'scripts/get_description.py'
       - 'docs/_ext/scylladb_metrics.py'
 
@@ -15,20 +15,20 @@ jobs:
   validate-metrics:
     runs-on: ubuntu-latest
     name: Check metrics documentation coverage
-
+    
     steps:
     - name: Checkout code
       uses: actions/checkout@v4
       with:
         submodules: true
-
+      
     - name: Set up Python
       uses: actions/setup-python@v6
       with:
         python-version: '3.10'
-
+        
     - name: Install dependencies
       run: pip install PyYAML
-
+        
     - name: Validate metrics
       run: python3 scripts/get_description.py --validate -c scripts/metrics-config.yml

.github/workflows/trigger-scylla-ci.yaml

@@ -3,13 +3,10 @@ name: Trigger Scylla CI Route
 on:
   issue_comment:
     types: [created]
-  pull_request_target:
-    types:
-      - unlabeled
 
 jobs:
   trigger-jenkins:
-    if: (github.event.comment.user.login != 'scylladbbot' && contains(github.event.comment.body, '@scylladbbot') && contains(github.event.comment.body, 'trigger-ci')) || github.event.label.name == 'conflicts'
+    if: github.event.comment.user.login != 'scylladbbot' && contains(github.event.comment.body, '@scylladbbot') && contains(github.event.comment.body, 'trigger-ci')
     runs-on: ubuntu-latest
     steps:
       - name: Trigger Scylla-CI-Route Jenkins Job

alternator/conditions.cc

@@ -42,7 +42,7 @@ comparison_operator_type get_comparison_operator(const rjson::value& comparison_
     if (!comparison_operator.IsString()) {
         throw api_error::validation(fmt::format("Invalid comparison operator definition {}", rjson::print(comparison_operator)));
     }
-    std::string op = rjson::to_string(comparison_operator);
+    std::string op = comparison_operator.GetString();
     auto it = ops.find(op);
     if (it == ops.end()) {
         throw api_error::validation(fmt::format("Unsupported comparison operator {}", op));
@@ -377,8 +377,8 @@ bool check_compare(const rjson::value* v1, const rjson::value& v2, const Compara
         return cmp(unwrap_number(*v1, cmp.diagnostic), unwrap_number(v2, cmp.diagnostic));
     }
     if (kv1.name == "S") {
-        return cmp(rjson::to_string_view(kv1.value),
-                   rjson::to_string_view(kv2.value));
+        return cmp(std::string_view(kv1.value.GetString(), kv1.value.GetStringLength()),
+                   std::string_view(kv2.value.GetString(), kv2.value.GetStringLength()));
     }
     if (kv1.name == "B") {
         auto d_kv1 = unwrap_bytes(kv1.value, v1_from_query);
@@ -470,9 +470,9 @@ static bool check_BETWEEN(const rjson::value* v, const rjson::value& lb, const r
         return check_BETWEEN(unwrap_number(*v, diag), unwrap_number(lb, diag), unwrap_number(ub, diag), bounds_from_query);
     }
     if (kv_v.name == "S") {
-        return check_BETWEEN(rjson::to_string_view(kv_v.value),
-                             rjson::to_string_view(kv_lb.value),
-                             rjson::to_string_view(kv_ub.value),
+        return check_BETWEEN(std::string_view(kv_v.value.GetString(), kv_v.value.GetStringLength()),
+                             std::string_view(kv_lb.value.GetString(), kv_lb.value.GetStringLength()),
+                             std::string_view(kv_ub.value.GetString(), kv_ub.value.GetStringLength()),
                              bounds_from_query);
     }
     if (kv_v.name == "B") {

alternator/consumed_capacity.cc

@@ -8,8 +8,6 @@
 
 #include "consumed_capacity.hh"
 #include "error.hh"
-#include "utils/rjson.hh"
-#include <fmt/format.h>
 
 namespace alternator {
 
@@ -34,12 +32,12 @@ bool consumed_capacity_counter::should_add_capacity(const rjson::value& request)
     if (!return_consumed->IsString()) {
         throw api_error::validation("Non-string ReturnConsumedCapacity field in request");
     }
-    std::string_view consumed = rjson::to_string_view(*return_consumed);
+    std::string consumed = return_consumed->GetString();
     if (consumed == "INDEXES") {
         throw api_error::validation("INDEXES consumed capacity is not supported");
     }
     if (consumed != "TOTAL") {
-        throw api_error::validation(fmt::format("Unknown consumed capacity {}", consumed));
+        throw api_error::validation("Unknown consumed capacity "+ consumed);
     }
     return true;
 }

alternator/controller.cc

@@ -169,7 +169,7 @@ future<> controller::request_stop_server() {
     });
 }
 
-future<utils::chunked_vector<foreign_ptr<std::unique_ptr<client_data>>>> controller::get_client_data() {
+future<utils::chunked_vector<client_data>> controller::get_client_data() {
     return _server.local().get_client_data();
 }
 

alternator/controller.hh

@@ -93,7 +93,7 @@ public:
     // This virtual function is called (on each shard separately) when the
     // virtual table "system.clients" is read. It is expected to generate a
     // list of clients connected to this server (on this shard).
-    virtual future<utils::chunked_vector<foreign_ptr<std::unique_ptr<client_data>>>> get_client_data() override;
+    virtual future<utils::chunked_vector<client_data>> get_client_data() override;
 };
 
 }

alternator/executor.cc

@@ -419,7 +419,7 @@ static std::optional<std::string> find_table_name(const rjson::value& request) {
     if (!table_name_value->IsString()) {
         throw api_error::validation("Non-string TableName field in request");
     }
-    std::string table_name = rjson::to_string(*table_name_value);
+    std::string table_name = table_name_value->GetString();
     return table_name;
 }
 
@@ -546,7 +546,7 @@ get_table_or_view(service::storage_proxy& proxy, const rjson::value& request) {
             // does exist but the index does not (ValidationException).
             if (proxy.data_dictionary().has_schema(keyspace_name, orig_table_name)) {
                 throw api_error::validation(
-                    fmt::format("Requested resource not found: Index '{}' for table '{}'", rjson::to_string_view(*index_name), orig_table_name));
+                    fmt::format("Requested resource not found: Index '{}' for table '{}'", index_name->GetString(), orig_table_name));
             } else {
                 throw api_error::resource_not_found(
                     fmt::format("Requested resource not found: Table: {} not found", orig_table_name));
@@ -587,7 +587,7 @@ static std::string get_string_attribute(const rjson::value& value, std::string_v
         throw api_error::validation(fmt::format("Expected string value for attribute {}, got: {}",
                 attribute_name, value));
     }
-    return rjson::to_string(*attribute_value);
+    return std::string(attribute_value->GetString(), attribute_value->GetStringLength());
 }
 
 // Convenience function for getting the value of a boolean attribute, or a
@@ -1080,8 +1080,8 @@ static void add_column(schema_builder& builder, const std::string& name, const r
     }
     for (auto it = attribute_definitions.Begin(); it != attribute_definitions.End(); ++it) {
         const rjson::value& attribute_info = *it;
-        if (rjson::to_string_view(attribute_info["AttributeName"]) == name) {
-            std::string_view type = rjson::to_string_view(attribute_info["AttributeType"]);
+        if (attribute_info["AttributeName"].GetString() == name) {
+            auto type = attribute_info["AttributeType"].GetString();
             data_type dt = parse_key_type(type);
             if (computed_column) {
                 // Computed column for GSI (doesn't choose a real column as-is
@@ -1116,7 +1116,7 @@ static std::pair<std::string, std::string> parse_key_schema(const rjson::value&
         throw api_error::validation("First element of KeySchema must be an object");
     }
     const rjson::value *v = rjson::find((*key_schema)[0], "KeyType");
-    if (!v || !v->IsString() || rjson::to_string_view(*v) != "HASH") {
+    if (!v || !v->IsString() || v->GetString() != std::string("HASH")) {
         throw api_error::validation("First key in KeySchema must be a HASH key");
     }
     v = rjson::find((*key_schema)[0], "AttributeName");
@@ -1124,14 +1124,14 @@ static std::pair<std::string, std::string> parse_key_schema(const rjson::value&
         throw api_error::validation("First key in KeySchema must have string AttributeName");
     }
     validate_attr_name_length(supplementary_context, v->GetStringLength(), true, "HASH key in KeySchema - ");
-    std::string hash_key = rjson::to_string(*v);
+    std::string hash_key = v->GetString();
     std::string range_key;
     if (key_schema->Size() == 2) {
         if (!(*key_schema)[1].IsObject()) {
             throw api_error::validation("Second element of KeySchema must be an object");
         }
         v = rjson::find((*key_schema)[1], "KeyType");
-        if (!v || !v->IsString() || rjson::to_string_view(*v) != "RANGE") {
+        if (!v || !v->IsString() || v->GetString() != std::string("RANGE")) {
             throw api_error::validation("Second key in KeySchema must be a RANGE key");
         }
         v = rjson::find((*key_schema)[1], "AttributeName");
@@ -1799,11 +1799,6 @@ static future<executor::request_return_type> create_table_on_shard0(service::cli
                 }
             }
         }
-        // Creating an index in tablets mode requires the rf_rack_valid_keyspaces option to be enabled.
-        // GSI and LSI indexes are based on materialized views which require this option to avoid consistency issues.
-        if (!view_builders.empty() && ksm->uses_tablets() && !sp.data_dictionary().get_config().rf_rack_valid_keyspaces()) {
-            co_return api_error::validation("GlobalSecondaryIndexes and LocalSecondaryIndexes with tablets require the rf_rack_valid_keyspaces option to be enabled.");
-        }
         try {
             schema_mutations = service::prepare_new_keyspace_announcement(sp.local_db(), ksm, ts);
         } catch (exceptions::already_exists_exception&) {
@@ -1892,8 +1887,8 @@ future<executor::request_return_type> executor::create_table(client_state& clien
         std::string def_type = type_to_string(def.type);
         for (auto it = attribute_definitions.Begin(); it != attribute_definitions.End(); ++it) {
             const rjson::value& attribute_info = *it;
-            if (rjson::to_string_view(attribute_info["AttributeName"]) == def.name_as_text()) {
-                std::string_view type = rjson::to_string_view(attribute_info["AttributeType"]);
+            if (attribute_info["AttributeName"].GetString() == def.name_as_text()) {
+                auto type = attribute_info["AttributeType"].GetString();
                 if (type != def_type) {
                     throw api_error::validation(fmt::format("AttributeDefinitions redefined {} to {} already a key attribute of type {} in this table", def.name_as_text(), type, def_type));
                 }
@@ -2024,10 +2019,6 @@ future<executor::request_return_type> executor::update_table(client_state& clien
                             co_return api_error::validation(fmt::format(
                                 "LSI {} already exists in table {}, can't use same name for GSI", index_name, table_name));
                         }
-                        if (p.local().local_db().find_keyspace(keyspace_name).get_replication_strategy().uses_tablets() &&
-                                !p.local().data_dictionary().get_config().rf_rack_valid_keyspaces()) {
-                            co_return api_error::validation("GlobalSecondaryIndexes with tablets require the rf_rack_valid_keyspaces option to be enabled.");
-                        }
 
                         elogger.trace("Adding GSI {}", index_name);
                         // FIXME: read and handle "Projection" parameter. This will
@@ -2232,12 +2223,12 @@ void validate_value(const rjson::value& v, const char* caller) {
 
 // The put_or_delete_item class builds the mutations needed by the PutItem and
 // DeleteItem operations - either as stand-alone commands or part of a list
-// of commands in BatchWriteItem.
+// of commands in BatchWriteItems.
 // put_or_delete_item splits each operation into two stages: Constructing the
 // object parses and validates the user input (throwing exceptions if there
 // are input errors). Later, build() generates the actual mutation, with a
 // specified timestamp. This split is needed because of the peculiar needs of
-// BatchWriteItem and LWT. BatchWriteItem needs all parsing to happen before
+// BatchWriteItems and LWT. BatchWriteItems needs all parsing to happen before
 // any writing happens (if one of the commands has an error, none of the
 // writes should be done). LWT makes it impossible for the parse step to
 // generate "mutation" objects, because the timestamp still isn't known.
@@ -2371,7 +2362,7 @@ put_or_delete_item::put_or_delete_item(const rjson::value& item, schema_ptr sche
     _cells = std::vector<cell>();
     _cells->reserve(item.MemberCount());
     for (auto it = item.MemberBegin(); it != item.MemberEnd(); ++it) {
-        bytes column_name = to_bytes(rjson::to_string_view(it->name));
+        bytes column_name = to_bytes(it->name.GetString());
         validate_value(it->value, "PutItem");
         const column_definition* cdef = find_attribute(*schema, column_name);
         validate_attr_name_length("", column_name.size(), cdef && cdef->is_primary_key());
@@ -2748,7 +2739,7 @@ future<executor::request_return_type> rmw_operation::execute(service::storage_pr
     auto read_command = needs_read_before_write ?
             previous_item_read_command(proxy, schema(), _ck, selection) :
             nullptr;
-    return proxy.cas(schema(), std::move(*cas_shard), *this, read_command, to_partition_ranges(*schema(), _pk),
+    return proxy.cas(schema(), std::move(*cas_shard), shared_from_this(), read_command, to_partition_ranges(*schema(), _pk),
             {timeout, std::move(permit), client_state, trace_state},
             db::consistency_level::LOCAL_SERIAL, db::consistency_level::LOCAL_QUORUM, timeout, timeout, true, std::move(cdc_opts)).then([this, read_command, &wcu_total] (bool is_applied) mutable {
         if (!is_applied) {
@@ -2792,10 +2783,10 @@ static void verify_all_are_used(const rjson::value* field,
         return;
     }
     for (auto it = field->MemberBegin(); it != field->MemberEnd(); ++it) {
-        if (!used.contains(rjson::to_string(it->name))) {
+        if (!used.contains(it->name.GetString())) {
             throw api_error::validation(
                 format("{} has spurious '{}', not used in {}",
-                    field_name, rjson::to_string_view(it->name), operation));
+                    field_name, it->name.GetString(), operation));
         }
     }
 }
@@ -3009,7 +3000,7 @@ future<executor::request_return_type> executor::delete_item(client_state& client
 }
 
 static schema_ptr get_table_from_batch_request(const service::storage_proxy& proxy, const rjson::value::ConstMemberIterator& batch_request) {
-    sstring table_name = rjson::to_sstring(batch_request->name); // JSON keys are always strings
+    sstring table_name = batch_request->name.GetString(); // JSON keys are always strings
     try {
         return proxy.data_dictionary().find_schema(sstring(executor::KEYSPACE_NAME_PREFIX) + table_name, table_name);
     } catch(data_dictionary::no_such_column_family&) {
@@ -3035,20 +3026,17 @@ struct primary_key_equal {
 };
 
 // This is a cas_request subclass for applying given put_or_delete_items to
-// one partition using LWT as part as BatchWriteItem. This is a write-only
+// one partition using LWT as part as BatchWriteItems. This is a write-only
 // operation, not needing the previous value of the item (the mutation to be
 // done is known prior to starting the operation). Nevertheless, we want to
 // do this mutation via LWT to ensure that it is serialized with other LWT
 // mutations to the same partition.
-// 
-// The std::vector<put_or_delete_item> must remain alive until the
-// storage_proxy::cas() future is resolved.
 class put_or_delete_item_cas_request : public service::cas_request {
     schema_ptr schema;
-    const std::vector<put_or_delete_item>& _mutation_builders;
+    std::vector<put_or_delete_item> _mutation_builders;
 public:
-    put_or_delete_item_cas_request(schema_ptr s, const std::vector<put_or_delete_item>& b) :
-        schema(std::move(s)), _mutation_builders(b) { }
+    put_or_delete_item_cas_request(schema_ptr s, std::vector<put_or_delete_item>&& b) :
+        schema(std::move(s)), _mutation_builders(std::move(b)) { }
     virtual ~put_or_delete_item_cas_request() = default;
     virtual std::optional<mutation> apply(foreign_ptr<lw_shared_ptr<query::result>> qr, const query::partition_slice& slice, api::timestamp_type ts, cdc::per_request_options& cdc_opts) override {
         std::optional<mutation> ret;
@@ -3064,48 +3052,20 @@ public:
     }
 };
 
-future<> executor::cas_write(schema_ptr schema, service::cas_shard cas_shard, const dht::decorated_key& dk,
-        const std::vector<put_or_delete_item>& mutation_builders, service::client_state& client_state,
-        tracing::trace_state_ptr trace_state, service_permit permit)
-{
-    if (!cas_shard.this_shard()) {
-        _stats.shard_bounce_for_lwt++;
-        return container().invoke_on(cas_shard.shard(), _ssg,
-                    [cs = client_state.move_to_other_shard(),
-                    &mb = mutation_builders,
-                    &dk,
-                    ks = schema->ks_name(),
-                    cf = schema->cf_name(),
-                    gt = tracing::global_trace_state_ptr(trace_state),
-                    permit = std::move(permit)]
-                    (executor& self) mutable {
-            return do_with(cs.get(), [&mb, &dk, ks = std::move(ks), cf = std::move(cf),
-                                    trace_state = tracing::trace_state_ptr(gt), &self]
-                                    (service::client_state& client_state) mutable {
-                auto schema = self._proxy.data_dictionary().find_schema(ks, cf);
-                service::cas_shard cas_shard(*schema, dk.token());
-
-                //FIXME: Instead of passing empty_service_permit() to the background operation,
-                // the current permit's lifetime should be prolonged, so that it's destructed
-                // only after all background operations are finished as well.
-                return self.cas_write(schema, std::move(cas_shard), dk, mb, client_state, std::move(trace_state), empty_service_permit());
-            });
-        });
-    }
-
+static future<> cas_write(service::storage_proxy& proxy, schema_ptr schema, service::cas_shard cas_shard, dht::decorated_key dk, std::vector<put_or_delete_item>&& mutation_builders,
+        service::client_state& client_state, tracing::trace_state_ptr trace_state, service_permit permit) {
     auto timeout = executor::default_timeout();
-    auto op = std::make_unique<put_or_delete_item_cas_request>(schema, mutation_builders);
-    auto* op_ptr = op.get();
+    auto op = seastar::make_shared<put_or_delete_item_cas_request>(schema, std::move(mutation_builders));
     auto cdc_opts = cdc::per_request_options{
         .alternator = true,
         .alternator_streams_increased_compatibility =
-                schema->cdc_options().enabled() && _proxy.data_dictionary().get_config().alternator_streams_increased_compatibility(),
+                schema->cdc_options().enabled() && proxy.data_dictionary().get_config().alternator_streams_increased_compatibility(),
     };
-    return _proxy.cas(schema, std::move(cas_shard), *op_ptr, nullptr, to_partition_ranges(dk),
+    return proxy.cas(schema, std::move(cas_shard), op, nullptr, to_partition_ranges(dk),
             {timeout, std::move(permit), client_state, trace_state},
             db::consistency_level::LOCAL_SERIAL, db::consistency_level::LOCAL_QUORUM,
-            timeout, timeout, true, std::move(cdc_opts)).finally([op = std::move(op)]{}).discard_result();
-    // We discarded cas()'s future value ("is_applied") because BatchWriteItem
+            timeout, timeout, true, std::move(cdc_opts)).discard_result();
+    // We discarded cas()'s future value ("is_applied") because BatchWriteItems
     // does not need to support conditional updates.
 }
 
@@ -3127,11 +3087,13 @@ struct schema_decorated_key_equal {
 
 // FIXME: if we failed writing some of the mutations, need to return a list
 // of these failed mutations rather than fail the whole write (issue #5650).
-future<> executor::do_batch_write(
+static future<> do_batch_write(service::storage_proxy& proxy,
+        smp_service_group ssg,
         std::vector<std::pair<schema_ptr, put_or_delete_item>> mutation_builders,
         service::client_state& client_state,
         tracing::trace_state_ptr trace_state,
-        service_permit permit) {
+        service_permit permit,
+        stats& stats) {
     if (mutation_builders.empty()) {
         return make_ready_future<>();
     }
@@ -3153,7 +3115,7 @@ future<> executor::do_batch_write(
             mutations.push_back(b.second.build(b.first, now));
             any_cdc_enabled |= b.first->cdc_options().enabled();
         }
-        return _proxy.mutate(std::move(mutations),
+        return proxy.mutate(std::move(mutations),
                 db::consistency_level::LOCAL_QUORUM,
                 executor::default_timeout(),
                 trace_state,
@@ -3162,48 +3124,55 @@ future<> executor::do_batch_write(
                 false,
                 cdc::per_request_options{
                     .alternator = true,
-                    .alternator_streams_increased_compatibility = any_cdc_enabled && _proxy.data_dictionary().get_config().alternator_streams_increased_compatibility(),
+                    .alternator_streams_increased_compatibility = any_cdc_enabled && proxy.data_dictionary().get_config().alternator_streams_increased_compatibility(),
                 });
     } else {
         // Do the write via LWT:
         // Multiple mutations may be destined for the same partition, adding
         // or deleting different items of one partition. Join them together
         // because we can do them in one cas() call.
-        using map_type = std::unordered_map<schema_decorated_key, 
-            std::vector<put_or_delete_item>, 
-            schema_decorated_key_hash, 
-            schema_decorated_key_equal>;
-        auto key_builders = std::make_unique<map_type>(1, schema_decorated_key_hash{}, schema_decorated_key_equal{});
-        for (auto&& b : std::move(mutation_builders)) {
-            auto [it, added] = key_builders->try_emplace(schema_decorated_key {
-                .schema = b.first,
-                .dk = dht::decorate_key(*b.first, b.second.pk())
-            });
+        std::unordered_map<schema_decorated_key, std::vector<put_or_delete_item>, schema_decorated_key_hash, schema_decorated_key_equal>
+            key_builders(1, schema_decorated_key_hash{}, schema_decorated_key_equal{});
+        for (auto& b : mutation_builders) {
+            auto dk = dht::decorate_key(*b.first, b.second.pk());
+            auto [it, added] = key_builders.try_emplace(schema_decorated_key{b.first, dk});
             it->second.push_back(std::move(b.second));
         }
-        auto* key_builders_ptr = key_builders.get();
-        return parallel_for_each(*key_builders_ptr, [this, &client_state, trace_state, permit = std::move(permit)] (const auto& e) {
-            _stats.write_using_lwt++;
+        return parallel_for_each(std::move(key_builders), [&proxy, &client_state, &stats, trace_state, ssg, permit = std::move(permit)] (auto& e) {
+            stats.write_using_lwt++;
             auto desired_shard = service::cas_shard(*e.first.schema, e.first.dk.token());
-            auto s = e.first.schema;
+            if (desired_shard.this_shard()) {
+                return cas_write(proxy, e.first.schema, std::move(desired_shard), e.first.dk, std::move(e.second), client_state, trace_state, permit);
+            } else {
+                stats.shard_bounce_for_lwt++;
+                return proxy.container().invoke_on(desired_shard.shard(), ssg,
+                            [cs = client_state.move_to_other_shard(),
+                             mb = e.second,
+                             dk = e.first.dk,
+                             ks = e.first.schema->ks_name(),
+                             cf = e.first.schema->cf_name(),
+                             gt =  tracing::global_trace_state_ptr(trace_state),
+                             permit = std::move(permit)]
+                            (service::storage_proxy& proxy) mutable {
+                    return do_with(cs.get(), [&proxy, mb = std::move(mb), dk = std::move(dk), ks = std::move(ks), cf = std::move(cf),
+                                              trace_state = tracing::trace_state_ptr(gt)]
+                                              (service::client_state& client_state) mutable {
+                        auto schema = proxy.data_dictionary().find_schema(ks, cf);
 
-            static const auto* injection_name = "alternator_executor_batch_write_wait";
-            return utils::get_local_injector().inject(injection_name, [s = std::move(s)] (auto& handler) -> future<> {
-                const auto ks = handler.get("keyspace");
-                const auto cf = handler.get("table");
-                const auto shard = std::atoll(handler.get("shard")->data());
-                if (ks == s->ks_name() && cf == s->cf_name() && shard == this_shard_id()) {
-                    elogger.info("{}: hit", injection_name);
-                    co_await handler.wait_for_message(std::chrono::steady_clock::now() + std::chrono::minutes{5});
-                    elogger.info("{}: continue", injection_name);
-                }
-            }).then([&e, desired_shard = std::move(desired_shard),
-                 &client_state, trace_state = std::move(trace_state), permit = std::move(permit), this]() mutable
-            {
-                return cas_write(e.first.schema, std::move(desired_shard), e.first.dk,
-                    std::move(e.second), client_state, std::move(trace_state), std::move(permit));
-            });
-        }).finally([key_builders = std::move(key_builders)]{});
+                        // The desired_shard on the original shard remains alive for the duration
+                        // of cas_write on this shard and prevents any tablet operations.
+                        // However, we need a local instance of cas_shard on this shard
+                        // to pass it to sp::cas, so we just create a new one.
+                        service::cas_shard cas_shard(*schema, dk.token());
+
+                        //FIXME: Instead of passing empty_service_permit() to the background operation,
+                        // the current permit's lifetime should be prolonged, so that it's destructed
+                        // only after all background operations are finished as well.
+                        return cas_write(proxy, schema, std::move(cas_shard), dk, std::move(mb), client_state, std::move(trace_state), empty_service_permit());
+                    });
+                }).finally([desired_shard = std::move(desired_shard)]{});
+            }
+        });
     }
 }
 
@@ -3350,7 +3319,7 @@ future<executor::request_return_type> executor::batch_write_item(client_state& c
     _stats.wcu_total[stats::DELETE_ITEM] += wcu_delete_units;
     _stats.api_operations.batch_write_item_batch_total += total_items;
     _stats.api_operations.batch_write_item_histogram.add(total_items);
-    co_await do_batch_write(std::move(mutation_builders), client_state, trace_state, std::move(permit));
+    co_await do_batch_write(_proxy, _ssg, std::move(mutation_builders), client_state, trace_state, std::move(permit), _stats);
     // FIXME: Issue #5650: If we failed writing some of the updates,
     // need to return a list of these failed updates in UnprocessedItems
     // rather than fail the whole write (issue #5650).
@@ -3395,7 +3364,7 @@ static bool hierarchy_filter(rjson::value& val, const attribute_path_map_node<T>
         }
         rjson::value newv = rjson::empty_object();
         for (auto it = v.MemberBegin(); it != v.MemberEnd(); ++it) {
-            std::string attr = rjson::to_string(it->name);
+            std::string attr = it->name.GetString();
             auto x = members.find(attr);
             if (x != members.end()) {
                 if (x->second) {
@@ -3615,7 +3584,7 @@ static std::optional<attrs_to_get> calculate_attrs_to_get(const rjson::value& re
         const rjson::value& attributes_to_get = req["AttributesToGet"];
         attrs_to_get ret;
         for (auto it = attributes_to_get.Begin(); it != attributes_to_get.End(); ++it) {
-            attribute_path_map_add("AttributesToGet", ret, rjson::to_string(*it));
+            attribute_path_map_add("AttributesToGet", ret, it->GetString());
             validate_attr_name_length("AttributesToGet", it->GetStringLength(), false);
         }
         if (ret.empty()) {
@@ -4281,12 +4250,12 @@ inline void update_item_operation::apply_attribute_updates(const std::unique_ptr
         attribute_collector& modified_attrs, bool& any_updates, bool& any_deletes) const {
     for (auto it = _attribute_updates->MemberBegin(); it != _attribute_updates->MemberEnd(); ++it) {
         // Note that it.key() is the name of the column, *it is the operation
-        bytes column_name = to_bytes(rjson::to_string_view(it->name));
+        bytes column_name = to_bytes(it->name.GetString());
         const column_definition* cdef = _schema->get_column_definition(column_name);
         if (cdef && cdef->is_primary_key()) {
-            throw api_error::validation(format("UpdateItem cannot update key column {}", rjson::to_string_view(it->name)));
+            throw api_error::validation(format("UpdateItem cannot update key column {}", it->name.GetString()));
         }
-        std::string action = rjson::to_string((it->value)["Action"]);
+        std::string action = (it->value)["Action"].GetString();
         if (action == "DELETE") {
             // The DELETE operation can do two unrelated tasks. Without a
             // "Value" option, it is used to delete an attribute. With a
@@ -5483,7 +5452,7 @@ calculate_bounds_conditions(schema_ptr schema, const rjson::value& conditions) {
     std::vector<query::clustering_range> ck_bounds;
 
     for (auto it = conditions.MemberBegin(); it != conditions.MemberEnd(); ++it) {
-        sstring key = rjson::to_sstring(it->name);
+        std::string key = it->name.GetString();
         const rjson::value& condition = it->value;
 
         const rjson::value& comp_definition = rjson::get(condition, "ComparisonOperator");
@@ -5491,13 +5460,13 @@ calculate_bounds_conditions(schema_ptr schema, const rjson::value& conditions) {
 
         const column_definition& pk_cdef = schema->partition_key_columns().front();
         const column_definition* ck_cdef = schema->clustering_key_size() > 0 ? &schema->clustering_key_columns().front() : nullptr;
-        if (key == pk_cdef.name_as_text()) {
+        if (sstring(key) == pk_cdef.name_as_text()) {
             if (!partition_ranges.empty()) {
                 throw api_error::validation("Currently only a single restriction per key is allowed");
             }
             partition_ranges.push_back(calculate_pk_bound(schema, pk_cdef, comp_definition, attr_list));
         }
-        if (ck_cdef && key == ck_cdef->name_as_text()) {
+        if (ck_cdef && sstring(key) == ck_cdef->name_as_text()) {
             if (!ck_bounds.empty()) {
                 throw api_error::validation("Currently only a single restriction per key is allowed");
             }
@@ -5898,7 +5867,7 @@ future<executor::request_return_type> executor::list_tables(client_state& client
 
     rjson::value* exclusive_start_json = rjson::find(request, "ExclusiveStartTableName");
     rjson::value* limit_json = rjson::find(request, "Limit");
-    std::string exclusive_start = exclusive_start_json ? rjson::to_string(*exclusive_start_json) : "";
+    std::string exclusive_start = exclusive_start_json ? exclusive_start_json->GetString() : "";
     int limit = limit_json ? limit_json->GetInt() : 100;
     if (limit < 1 || limit > 100) {
         co_return api_error::validation("Limit must be greater than 0 and no greater than 100");

alternator/executor.hh

@@ -40,7 +40,6 @@ namespace cql3::selection {
 
 namespace service {
     class storage_proxy;
-    class cas_shard;
 }
 
 namespace cdc {
@@ -58,7 +57,6 @@ class schema_builder;
 namespace alternator {
 
 class rmw_operation;
-class put_or_delete_item;
 
 schema_ptr get_table(service::storage_proxy& proxy, const rjson::value& request);
 bool is_alternator_keyspace(const sstring& ks_name);
@@ -221,16 +219,6 @@ private:
 
     static void describe_key_schema(rjson::value& parent, const schema&, std::unordered_map<std::string,std::string> * = nullptr, const std::map<sstring, sstring> *tags = nullptr);
 
-    future<> do_batch_write(
-        std::vector<std::pair<schema_ptr, put_or_delete_item>> mutation_builders,
-        service::client_state& client_state,
-        tracing::trace_state_ptr trace_state,
-        service_permit permit);
-
-    future<> cas_write(schema_ptr schema, service::cas_shard cas_shard, const dht::decorated_key& dk,
-        const std::vector<put_or_delete_item>& mutation_builders, service::client_state& client_state,
-        tracing::trace_state_ptr trace_state, service_permit permit);
-
 public:
     static void describe_key_schema(rjson::value& parent, const schema& schema, std::unordered_map<std::string,std::string>&, const std::map<sstring, sstring> *tags = nullptr);
 

alternator/serialization.cc

@@ -496,7 +496,7 @@ const std::pair<std::string, const rjson::value*> unwrap_set(const rjson::value&
         return {"", nullptr};
     }
     auto it = v.MemberBegin();
-    const std::string it_key = rjson::to_string(it->name);
+    const std::string it_key = it->name.GetString();
     if (it_key != "SS" && it_key != "BS" && it_key != "NS") {
         return {std::move(it_key), nullptr};
     }

alternator/server.cc

@@ -708,12 +708,8 @@ future<executor::request_return_type> server::handle_api_request(std::unique_ptr
     // As long as the system_clients_entry object is alive, this request will
     // be visible in the "system.clients" virtual table. When requested, this
     // entry will be formatted by server::ongoing_request::make_client_data().
-    auto user_agent_header = co_await _connection_options_keys_and_values.get_or_load(req->get_header("User-Agent"), [] (const client_options_cache_key_type&) {
-        return make_ready_future<options_cache_value_type>(options_cache_value_type{});
-    });
-
     auto system_clients_entry = _ongoing_requests.emplace(
-        req->get_client_address(), std::move(user_agent_header),
+        req->get_client_address(), req->get_header("User-Agent"),
         username, current_scheduling_group(),
         req->get_protocol_name() == "https");
 
@@ -989,10 +985,10 @@ client_data server::ongoing_request::make_client_data() const {
     return cd;
 }
 
-future<utils::chunked_vector<foreign_ptr<std::unique_ptr<client_data>>>> server::get_client_data() {
-    utils::chunked_vector<foreign_ptr<std::unique_ptr<client_data>>> ret;
+future<utils::chunked_vector<client_data>> server::get_client_data() {
+    utils::chunked_vector<client_data> ret;
     co_await _ongoing_requests.for_each_gently([&ret] (const ongoing_request& r) {
-        ret.emplace_back(make_foreign(std::make_unique<client_data>(r.make_client_data())));
+        ret.emplace_back(r.make_client_data());
     });
     co_return ret;
 }

alternator/server.hh

@@ -55,7 +55,6 @@ class server : public peering_sharded_service<server> {
     // though it isn't really relevant for Alternator which defines its own
     // timeouts separately. We can create this object only once.
     updateable_timeout_config _timeout_config;
-    client_options_cache_type _connection_options_keys_and_values;
 
     alternator_callbacks_map _callbacks;
 
@@ -89,7 +88,7 @@ class server : public peering_sharded_service<server> {
     // is called when reading the "system.clients" virtual table.
     struct ongoing_request {
         socket_address _client_address;
-        client_options_cache_entry_type _user_agent;
+        sstring _user_agent;
         sstring _username;
         scheduling_group _scheduling_group;
         bool _is_https;
@@ -108,7 +107,7 @@ public:
     // table "system.clients" is read. It is expected to generate a list of
     // clients connected to this server (on this shard). This function is
     // called by alternator::controller::get_client_data().
-    future<utils::chunked_vector<foreign_ptr<std::unique_ptr<client_data>>>> get_client_data();
+    future<utils::chunked_vector<client_data>> get_client_data();
 private:
     void set_routes(seastar::httpd::routes& r);
     // If verification succeeds, returns the authenticated user's username

alternator/ttl.cc

@@ -93,7 +93,7 @@ future<executor::request_return_type> executor::update_time_to_live(client_state
     if (v->GetStringLength() < 1 || v->GetStringLength() > 255) {
         co_return api_error::validation("The length of AttributeName must be between 1 and 255");
     }
-    sstring attribute_name = rjson::to_sstring(*v);
+    sstring attribute_name(v->GetString(), v->GetStringLength());
 
     co_await verify_permission(_enforce_authorization, _warn_authorization, client_state, schema, auth::permission::ALTER, _stats);
     co_await db::modify_tags(_mm, schema->ks_name(), schema->cf_name(), [&](std::map<sstring, sstring>& tags_map) {

api/CMakeLists.txt

@@ -31,7 +31,6 @@ set(swagger_files
   api-doc/column_family.json
   api-doc/commitlog.json
   api-doc/compaction_manager.json
-  api-doc/client_routes.json
   api-doc/config.json
   api-doc/cql_server_test.json
   api-doc/endpoint_snitch_info.json
@@ -69,7 +68,6 @@ target_sources(api
   PRIVATE
     api.cc
     cache_service.cc
-    client_routes.cc
     collectd.cc
     column_family.cc
     commitlog.cc

api/api-doc/client_routes.def.json

@@ -1,23 +0,0 @@
-    , "client_routes_entry": {
-        "id": "client_routes_entry",
-        "summary": "An entry storing client routes",
-        "properties": {
-            "connection_id": {"type": "string"},
-            "host_id": {"type": "string", "format": "uuid"},
-            "address": {"type": "string"},
-            "port": {"type": "integer"},
-            "tls_port": {"type": "integer"},
-            "alternator_port": {"type": "integer"},
-            "alternator_https_port": {"type": "integer"}
-        },
-        "required": ["connection_id", "host_id", "address"]
-    }
-    , "client_routes_key": {
-        "id": "client_routes_key",
-        "summary": "A key of client_routes_entry",
-        "properties": {
-            "connection_id": {"type": "string"},
-            "host_id": {"type": "string", "format": "uuid"}
-        }
-    }
-

api/api-doc/client_routes.json

@@ -1,74 +0,0 @@
-    , "/v2/client-routes":{
-        "get": {
-            "description":"List all client route entries",
-            "operationId":"get_client_routes",
-            "tags":["client_routes"],
-            "produces":[
-                "application/json"
-            ],
-            "parameters":[],
-            "responses":{
-                "200":{
-                    "schema":{
-                        "type":"array",
-                        "items":{ "$ref":"#/definitions/client_routes_entry" }
-                    }
-                },
-                "default":{
-                    "description":"unexpected error",
-                    "schema":{"$ref":"#/definitions/ErrorModel"}
-                }
-            }
-        },
-        "post": {
-            "description":"Upsert one or more client route entries",
-            "operationId":"set_client_routes",
-            "tags":["client_routes"],
-            "parameters":[
-                {
-                    "name":"body",
-                    "in":"body",
-                    "required":true,
-                    "schema":{
-                        "type":"array",
-                        "items":{ "$ref":"#/definitions/client_routes_entry" }
-                    }
-                }
-            ],
-            "responses":{
-                "200":{ "description": "OK" },
-                "default":{
-                    "description":"unexpected error",
-                    "schema":{ "$ref":"#/definitions/ErrorModel" }
-                }
-            }
-        },
-        "delete": {
-            "description":"Delete one or more client route entries",
-            "operationId":"delete_client_routes",
-            "tags":["client_routes"],
-            "parameters":[
-                {
-                    "name":"body",
-                    "in":"body",
-                    "required":true,
-                    "schema":{
-                        "type":"array",
-                        "items":{ "$ref":"#/definitions/client_routes_key" }
-                    }
-                }
-            ],
-            "responses":{
-                "200":{
-                    "description": "OK"
-                },
-                "default":{
-                    "description":"unexpected error",
-                    "schema":{
-                        "$ref":"#/definitions/ErrorModel"
-                    }
-                }
-            }
-        }
-    }
-

api/api-doc/storage_service.json

@@ -3051,7 +3051,7 @@
                   },
                   {
                      "name":"incremental_mode",
-                     "description":"Set the incremental repair mode. Can be 'disabled', 'incremental', or 'full'. 'incremental': The incremental repair logic is enabled. Unrepaired sstables will be included for repair. Repaired sstables will be skipped. The incremental repair states will be updated after repair. 'full': The incremental repair logic is enabled. Both repaired and unrepaired sstables will be included for repair. The incremental repair states will be updated after repair. 'disabled': The incremental repair logic is disabled completely. The incremental repair states, e.g., repaired_at in sstables and sstables_repaired_at in the system.tablets table, will not be updated after repair. When the option is not provided, it defaults to 'disabled' mode.",
+                     "description":"Set the incremental repair mode. Can be 'disabled', 'incremental', or 'full'. 'incremental': The incremental repair logic is enabled. Unrepaired sstables will be included for repair. Repaired sstables will be skipped. The incremental repair states will be updated after repair. 'full': The incremental repair logic is enabled. Both repaired and unrepaired sstables will be included for repair. The incremental repair states will be updated after repair. 'disabled': The incremental repair logic is disabled completely. The incremental repair states, e.g., repaired_at in sstables and sstables_repaired_at in the system.tablets table, will not be updated after repair. When the option is not provided, it defaults to incremental mode.",
                      "required":false,
                      "allowMultiple":false,
                      "type":"string",

api/api.cc

@@ -37,7 +37,6 @@
 #include "raft.hh"
 #include "gms/gossip_address_map.hh"
 #include "service_levels.hh"
-#include "client_routes.hh"
 
 logging::logger apilog("api");
 
@@ -68,11 +67,9 @@ future<> set_server_init(http_context& ctx) {
         rb02->set_api_doc(r);
         rb02->register_api_file(r, "swagger20_header");
         rb02->register_api_file(r, "metrics");
-        rb02->register_api_file(r, "client_routes");
         rb->register_function(r, "system",
                 "The system related API");
         rb02->add_definitions_file(r, "metrics");
-        rb02->add_definitions_file(r, "client_routes");
         set_system(ctx, r);
         rb->register_function(r, "error_injection",
             "The error injection API");
@@ -132,16 +129,6 @@ future<> unset_server_storage_service(http_context& ctx) {
     return ctx.http_server.set_routes([&ctx] (routes& r) { unset_storage_service(ctx, r); });
 }
 
-future<> set_server_client_routes(http_context& ctx, sharded<service::client_routes_service>& cr) {
-    return ctx.http_server.set_routes([&ctx, &cr] (routes& r) {
-        set_client_routes(ctx, r, cr);
-    });
-}
-
-future<> unset_server_client_routes(http_context& ctx) {
-    return ctx.http_server.set_routes([&ctx] (routes& r) { unset_client_routes(ctx, r); });
-}
-
 future<> set_load_meter(http_context& ctx, service::load_meter& lm) {
     return ctx.http_server.set_routes([&ctx, &lm] (routes& r) { set_load_meter(ctx, r, lm); });
 }

api/api_init.hh

@@ -29,7 +29,6 @@ class storage_proxy;
 class storage_service;
 class raft_group0_client;
 class raft_group_registry;
-class client_routes_service;
 
 } // namespace service
 
@@ -100,8 +99,6 @@ future<> set_server_snitch(http_context& ctx, sharded<locator::snitch_ptr>& snit
 future<> unset_server_snitch(http_context& ctx);
 future<> set_server_storage_service(http_context& ctx, sharded<service::storage_service>& ss, service::raft_group0_client&);
 future<> unset_server_storage_service(http_context& ctx);
-future<> set_server_client_routes(http_context& ctx, sharded<service::client_routes_service>& cr);
-future<> unset_server_client_routes(http_context& ctx);
 future<> set_server_sstables_loader(http_context& ctx, sharded<sstables_loader>& sst_loader);
 future<> unset_server_sstables_loader(http_context& ctx);
 future<> set_server_view_builder(http_context& ctx, sharded<db::view::view_builder>& vb, sharded<gms::gossiper>& g);

api/client_routes.cc

@@ -1,176 +0,0 @@
-/*
- * Copyright (C) 2025-present ScyllaDB
- *
- */
-
-/*
- * SPDX-License-Identifier: LicenseRef-ScyllaDB-Source-Available-1.0
- */
-
- #include <seastar/http/short_streams.hh>
-
-#include "client_routes.hh"
-#include "api/api.hh"
-#include "service/storage_service.hh"
-#include "service/client_routes.hh"
-#include "utils/rjson.hh"
-
-
-#include "api/api-doc/client_routes.json.hh"
-
-using namespace seastar::httpd;
-using namespace std::chrono_literals;
-using namespace json;
-
-extern logging::logger apilog;
-
-namespace api {
-
-static void validate_client_routes_endpoint(sharded<service::client_routes_service>& cr, sstring endpoint_name) {
-    if (!cr.local().get_feature_service().client_routes) {
-        apilog.warn("{}: called before the cluster feature was enabled", endpoint_name);
-        throw std::runtime_error(fmt::format("{} requires all nodes to support the CLIENT_ROUTES cluster feature", endpoint_name));
-    }
-}
-
-static sstring parse_string(const char* name, rapidjson::Value const& v) {
-    const auto it = v.FindMember(name);
-    if (it == v.MemberEnd()) {
-        throw bad_param_exception(fmt::format("Missing '{}'", name));
-    }
-    if (!it->value.IsString()) {
-        throw bad_param_exception(fmt::format("'{}' must be a string", name));
-    }
-    return {it->value.GetString(), it->value.GetStringLength()};
-}
-
-static std::optional<uint32_t> parse_port(const char* name, rapidjson::Value const& v) {
-    const auto it = v.FindMember(name);
-    if (it == v.MemberEnd()) {
-        return std::nullopt;
-    }
-    if (!it->value.IsInt()) {
-        throw bad_param_exception(fmt::format("'{}' must be an integer", name));
-    }
-    auto port = it->value.GetInt();
-    if (port < 1 || port > 65535) {
-        throw bad_param_exception(fmt::format("'{}' value={} is outside the allowed port range", name, port));
-    }
-    return port;
-}
-
-static std::vector<service::client_routes_service::client_route_entry> parse_set_client_array(const rapidjson::Document& root) {
-    if (!root.IsArray()) {
-        throw bad_param_exception("Body must be a JSON array");
-    }
-
-    std::vector<service::client_routes_service::client_route_entry> v;
-    v.reserve(root.GetArray().Size());
-    for (const auto& element : root.GetArray()) {
-        if (!element.IsObject()) { throw bad_param_exception("Each element must be object"); }
-
-        const auto port = parse_port("port", element);
-        const auto tls_port = parse_port("tls_port", element);
-        const auto alternator_port = parse_port("alternator_port", element);
-        const auto alternator_https_port = parse_port("alternator_https_port", element);
-
-        if (!port.has_value() && !tls_port.has_value() && !alternator_port.has_value() && !alternator_https_port.has_value()) {
-            throw bad_param_exception("At least one port field ('port', 'tls_port', 'alternator_port', 'alternator_https_port') must be specified");
-        }
-
-        v.emplace_back(
-            parse_string("connection_id", element),
-            utils::UUID{parse_string("host_id", element)},
-            parse_string("address", element),
-            port,
-            tls_port,
-            alternator_port,
-            alternator_https_port
-        );
-    }
-
-    return v;
-}
-
-static
-future<json::json_return_type>
-rest_set_client_routes(http_context& ctx, sharded<service::client_routes_service>& cr, std::unique_ptr<http::request> req) {
-    validate_client_routes_endpoint(cr, "rest_set_client_routes");
-
-    rapidjson::Document root;
-    auto content = co_await util::read_entire_stream_contiguous(*req->content_stream);
-    root.Parse(content.c_str());
-
-    co_await cr.local().set_client_routes(parse_set_client_array(root));
-    co_return seastar::json::json_void();
-}
-
-static std::vector<service::client_routes_service::client_route_key> parse_delete_client_array(const rapidjson::Document& root) {
-    if (!root.IsArray()) {
-        throw bad_param_exception("Body must be a JSON array");
-    }
-
-    std::vector<service::client_routes_service::client_route_key> v;
-    v.reserve(root.GetArray().Size());
-    for (const auto& element : root.GetArray()) {
-        v.emplace_back(
-            parse_string("connection_id", element),
-            utils::UUID{parse_string("host_id", element)}
-        );
-    }
-
-    return v;
-}
-
-static
-future<json::json_return_type>
-rest_delete_client_routes(http_context& ctx, sharded<service::client_routes_service>& cr, std::unique_ptr<http::request> req) {
-    validate_client_routes_endpoint(cr, "delete_client_routes");
-
-    rapidjson::Document root;
-    auto content = co_await util::read_entire_stream_contiguous(*req->content_stream);
-    root.Parse(content.c_str());
-
-    co_await cr.local().delete_client_routes(parse_delete_client_array(root));
-    co_return seastar::json::json_void();
-}
-
-static
-future<json::json_return_type>
-rest_get_client_routes(http_context& ctx, sharded<service::client_routes_service>& cr, std::unique_ptr<http::request> req) {
-    validate_client_routes_endpoint(cr, "get_client_routes");
-
-    co_return co_await cr.invoke_on(0, [] (service::client_routes_service& cr) -> future<json::json_return_type> {
-        co_return json::json_return_type(stream_range_as_array(co_await cr.get_client_routes(), [](const service::client_routes_service::client_route_entry & entry) {
-            seastar::httpd::client_routes_json::client_routes_entry obj;
-            obj.connection_id = entry.connection_id;
-            obj.host_id = fmt::to_string(entry.host_id);
-            obj.address = entry.address;
-            if (entry.port.has_value()) { obj.port = entry.port.value(); }
-            if (entry.tls_port.has_value()) { obj.tls_port = entry.tls_port.value(); }
-            if (entry.alternator_port.has_value()) { obj.alternator_port = entry.alternator_port.value(); }
-            if (entry.alternator_https_port.has_value()) { obj.alternator_https_port = entry.alternator_https_port.value(); }
-            return obj;
-        }));
-    });
-}
-
-void set_client_routes(http_context& ctx, routes& r, sharded<service::client_routes_service>& cr) {
-    seastar::httpd::client_routes_json::set_client_routes.set(r, [&ctx, &cr] (std::unique_ptr<seastar::http::request> req) {
-        return rest_set_client_routes(ctx, cr, std::move(req));
-    });
-    seastar::httpd::client_routes_json::delete_client_routes.set(r, [&ctx, &cr] (std::unique_ptr<seastar::http::request> req) {
-        return rest_delete_client_routes(ctx, cr, std::move(req));
-    });
-    seastar::httpd::client_routes_json::get_client_routes.set(r, [&ctx, &cr] (std::unique_ptr<seastar::http::request> req) {
-        return rest_get_client_routes(ctx, cr, std::move(req));
-    });
-}
-
-void unset_client_routes(http_context& ctx, routes& r) {
-    seastar::httpd::client_routes_json::set_client_routes.unset(r);
-    seastar::httpd::client_routes_json::delete_client_routes.unset(r);
-    seastar::httpd::client_routes_json::get_client_routes.unset(r);
-}
-
-}

api/client_routes.hh

@@ -1,20 +0,0 @@
-/*
- * Copyright (C) 2025-present ScyllaDB
- *
- */
-
-/*
- * SPDX-License-Identifier: LicenseRef-ScyllaDB-Source-Available-1.0
- */
-#pragma once
-
-#include <seastar/core/sharded.hh>
-#include <seastar/json/json_elements.hh>
-#include "api/api_init.hh"
-
-namespace api {
-
-void set_client_routes(http_context& ctx, httpd::routes& r, sharded<service::client_routes_service>& cr);
-void unset_client_routes(http_context& ctx, httpd::routes& r);
-
-}

auth/allow_all_authenticator.cc

@@ -9,6 +9,7 @@
 #include "auth/allow_all_authenticator.hh"
 
 #include "service/migration_manager.hh"
+#include "utils/alien_worker.hh"
 #include "utils/class_registrator.hh"
 
 namespace auth {
@@ -22,6 +23,7 @@ static const class_registrator<
         cql3::query_processor&,
         ::service::raft_group0_client&,
         ::service::migration_manager&,
-        cache&> registration("org.apache.cassandra.auth.AllowAllAuthenticator");
+        cache&,
+        utils::alien_worker&> registration("org.apache.cassandra.auth.AllowAllAuthenticator");
 
 }

auth/allow_all_authenticator.hh

@@ -14,6 +14,7 @@
 #include "auth/authenticator.hh"
 #include "auth/cache.hh"
 #include "auth/common.hh"
+#include "utils/alien_worker.hh"
 
 namespace cql3 {
 class query_processor;
@@ -29,7 +30,7 @@ extern const std::string_view allow_all_authenticator_name;
 
 class allow_all_authenticator final : public authenticator {
 public:
-    allow_all_authenticator(cql3::query_processor&, ::service::raft_group0_client&, ::service::migration_manager&, cache&) {
+    allow_all_authenticator(cql3::query_processor&, ::service::raft_group0_client&, ::service::migration_manager&, cache&, utils::alien_worker&) {
     }
 
     virtual future<> start() override {

auth/cache.cc

@@ -100,12 +100,10 @@ future<lw_shared_ptr<cache::role_record>> cache::fetch_role(const role_name_t& r
 }
 
 future<> cache::prune_all() noexcept {
-    for (auto it = _roles.begin(); it != _roles.end(); ) {
+    for (auto it = _roles.begin(); it != _roles.end(); it++) {
         if (it->second->version != _current_version) {
-            _roles.erase(it++);
+            _roles.erase(it);
             co_await coroutine::maybe_yield();
-        } else {
-            ++it;
         }
     }
     co_return;

auth/certificate_authenticator.cc

@@ -8,7 +8,6 @@
  */
 
 #include "auth/certificate_authenticator.hh"
-#include "auth/cache.hh"
 
 #include <boost/regex.hpp>
 #include <fmt/ranges.h>
@@ -35,13 +34,13 @@ static const class_registrator<auth::authenticator
     , cql3::query_processor&
     , ::service::raft_group0_client&
     , ::service::migration_manager&
-    , auth::cache&> cert_auth_reg(CERT_AUTH_NAME);
+    , utils::alien_worker&> cert_auth_reg(CERT_AUTH_NAME);
 
 enum class auth::certificate_authenticator::query_source {
     subject, altname
 };
 
-auth::certificate_authenticator::certificate_authenticator(cql3::query_processor& qp, ::service::raft_group0_client&, ::service::migration_manager&, auth::cache&)
+auth::certificate_authenticator::certificate_authenticator(cql3::query_processor& qp, ::service::raft_group0_client&, ::service::migration_manager&, utils::alien_worker&)
     : _queries([&] {
         auto& conf = qp.db().get_config();
         auto queries = conf.auth_certificate_role_queries();
@@ -76,9 +75,9 @@ auth::certificate_authenticator::certificate_authenticator(cql3::query_processor
                         throw std::invalid_argument(fmt::format("Invalid source: {}", map.at(cfg_source_attr)));
                     }
                     continue;
-                } catch (const std::out_of_range&) {
+                } catch (std::out_of_range&) {
                     // just fallthrough
-                } catch (const boost::regex_error&) {
+                } catch (boost::regex_error&) {
                     std::throw_with_nested(std::invalid_argument(fmt::format("Invalid query expression: {}", map.at(cfg_query_attr))));
                 }
             }

auth/certificate_authenticator.hh

@@ -10,6 +10,7 @@
 #pragma once
 
 #include "auth/authenticator.hh"
+#include "utils/alien_worker.hh"
 #include <boost/regex_fwd.hpp>  // IWYU pragma: keep
 
 namespace cql3 {
@@ -25,15 +26,13 @@ class raft_group0_client;
 
 namespace auth {
 
-class cache;
-
 extern const std::string_view certificate_authenticator_name;
 
 class certificate_authenticator : public authenticator {
     enum class query_source;
     std::vector<std::pair<query_source, boost::regex>> _queries;
 public:
-    certificate_authenticator(cql3::query_processor&, ::service::raft_group0_client&, ::service::migration_manager&, cache&);
+    certificate_authenticator(cql3::query_processor&, ::service::raft_group0_client&, ::service::migration_manager&, utils::alien_worker&);
     ~certificate_authenticator();
 
     future<> start() override;

auth/common.cc

@@ -94,7 +94,7 @@ static future<> create_legacy_metadata_table_if_missing_impl(
         try {
             co_return co_await mm.announce(co_await ::service::prepare_new_column_family_announcement(qp.proxy(), table, ts),
                     std::move(group0_guard), format("auth: create {} metadata table", table->cf_name()));
-        } catch (const exceptions::already_exists_exception&) {}
+        } catch (exceptions::already_exists_exception&) {}
     }
 }
 

auth/default_authorizer.cc

@@ -256,7 +256,7 @@ future<> default_authorizer::revoke_all(std::string_view role_name, ::service::g
         } else {
             co_await collect_mutations(_qp, mc, query, {sstring(role_name)});
         }
-    } catch (const exceptions::request_execution_exception& e) {
+    } catch (exceptions::request_execution_exception& e) {
         alogger.warn("CassandraAuthorizer failed to revoke all permissions of {}: {}", role_name, e);
     }
 }
@@ -293,13 +293,13 @@ future<> default_authorizer::revoke_all_legacy(const resource& resource) {
                                 [resource](auto ep) {
                     try {
                         std::rethrow_exception(ep);
-                    } catch (const exceptions::request_execution_exception& e) {
+                    } catch (exceptions::request_execution_exception& e) {
                         alogger.warn("CassandraAuthorizer failed to revoke all permissions on {}: {}", resource, e);
                     }
 
                 });
             });
-        } catch (const exceptions::request_execution_exception& e) {
+        } catch (exceptions::request_execution_exception& e) {
             alogger.warn("CassandraAuthorizer failed to revoke all permissions on {}: {}", resource, e);
             return make_ready_future();
         }

auth/password_authenticator.cc

@@ -49,7 +49,8 @@ static const class_registrator<
         cql3::query_processor&,
         ::service::raft_group0_client&,
         ::service::migration_manager&,
-        cache&> password_auth_reg("org.apache.cassandra.auth.PasswordAuthenticator");
+        cache&,
+        utils::alien_worker&> password_auth_reg("org.apache.cassandra.auth.PasswordAuthenticator");
 
 static thread_local auto rng_for_salt = std::default_random_engine(std::random_device{}());
 
@@ -63,13 +64,14 @@ std::string password_authenticator::default_superuser(const db::config& cfg) {
 password_authenticator::~password_authenticator() {
 }
 
-password_authenticator::password_authenticator(cql3::query_processor& qp, ::service::raft_group0_client& g0, ::service::migration_manager& mm, cache& cache)
+password_authenticator::password_authenticator(cql3::query_processor& qp, ::service::raft_group0_client& g0, ::service::migration_manager& mm, cache& cache, utils::alien_worker& hashing_worker)
     : _qp(qp)
     , _group0_client(g0)
     , _migration_manager(mm)
     , _cache(cache)
     , _stopped(make_ready_future<>()) 
     , _superuser(default_superuser(qp.db().get_config()))
+    , _hashing_worker(hashing_worker)
 {}
 
 static bool has_salted_hash(const cql3::untyped_result_set_row& row) {
@@ -328,18 +330,20 @@ future<authenticated_user> password_authenticator::authenticate(
             }
             salted_hash = role->salted_hash;
         }
-        const bool password_match = co_await passwords::check(password, *salted_hash);
+        const bool password_match = co_await _hashing_worker.submit<bool>([password = std::move(password), salted_hash] {
+            return passwords::check(password, *salted_hash);
+        });
         if (!password_match) {
             throw exceptions::authentication_exception("Username and/or password are incorrect");
         }
         co_return username;
-    } catch (const std::system_error &) {
+    } catch (std::system_error &) {
         std::throw_with_nested(exceptions::authentication_exception("Could not verify password"));
-    } catch (const exceptions::request_execution_exception& e) {
+    } catch (exceptions::request_execution_exception& e) {
         std::throw_with_nested(exceptions::authentication_exception(e.what()));
-    } catch (const exceptions::authentication_exception& e) {
+    } catch (exceptions::authentication_exception& e) {
         std::throw_with_nested(e);
-    } catch (const exceptions::unavailable_exception& e) {
+    } catch (exceptions::unavailable_exception& e) {
         std::throw_with_nested(exceptions::authentication_exception(e.get_message()));
     } catch (...) {
         std::throw_with_nested(exceptions::authentication_exception("authentication failed"));

auth/password_authenticator.hh

@@ -18,6 +18,7 @@
 #include "auth/passwords.hh"
 #include "auth/cache.hh"
 #include "service/raft/raft_group0_client.hh"
+#include "utils/alien_worker.hh"
 
 namespace db {
     class config;
@@ -48,12 +49,13 @@ class password_authenticator : public authenticator {
     shared_promise<> _superuser_created_promise;
     // We used to also support bcrypt, SHA-256, and MD5 (ref. scylladb#24524).
     constexpr static auth::passwords::scheme _scheme = passwords::scheme::sha_512;
+    utils::alien_worker& _hashing_worker;
 
 public:
     static db::consistency_level consistency_for_user(std::string_view role_name);
     static std::string default_superuser(const db::config&);
 
-    password_authenticator(cql3::query_processor&, ::service::raft_group0_client&, ::service::migration_manager&, cache&);
+    password_authenticator(cql3::query_processor&, ::service::raft_group0_client&, ::service::migration_manager&, cache&, utils::alien_worker&);
 
     ~password_authenticator();
 

auth/passwords.cc

@@ -7,8 +7,6 @@
  */
 
 #include "auth/passwords.hh"
-#include "utils/crypt_sha512.hh"
-#include <seastar/core/coroutine.hh>
 
 #include <cerrno>
 
@@ -23,46 +21,25 @@ static thread_local crypt_data tlcrypt = {};
 
 namespace detail {
 
-void verify_hashing_output(const char * res) {
-    if (!res || (res[0] == '*')) {
-        throw std::system_error(errno, std::system_category());
-    }
-}
-
 void verify_scheme(scheme scheme) {
     const sstring random_part_of_salt = "aaaabbbbccccdddd";
 
     const sstring salt = sstring(prefix_for_scheme(scheme)) + random_part_of_salt;
     const char* e = crypt_r("fisk", salt.c_str(), &tlcrypt);
-    try {
-        verify_hashing_output(e);
-    } catch (const std::system_error& ex) {
-        throw no_supported_schemes();
+
+    if (e && (e[0] != '*')) {
+        return;
     }
+
+    throw no_supported_schemes();
 }
 
 sstring hash_with_salt(const sstring& pass, const sstring& salt) {
     auto res = crypt_r(pass.c_str(), salt.c_str(), &tlcrypt);
-    verify_hashing_output(res);
-    return res;
-}
-
-seastar::future<sstring> hash_with_salt_async(const sstring& pass, const sstring& salt) {
-    sstring res;
-    // Only SHA-512 hashes for passphrases shorter than 256 bytes can be computed using
-    // the __crypt_sha512 method. For other computations, we fall back to the
-    // crypt_r implementation from `<crypt.h>`, which can stall.
-    if (salt.starts_with(prefix_for_scheme(scheme::sha_512)) && pass.size() <= 255) {
-        char buf[128];
-        const char * output_ptr = co_await __crypt_sha512(pass.c_str(), salt.c_str(), buf);
-        verify_hashing_output(output_ptr);
-        res = output_ptr;
-    } else {
-        const char * output_ptr = crypt_r(pass.c_str(), salt.c_str(), &tlcrypt);
-        verify_hashing_output(output_ptr);
-        res = output_ptr;
+    if (!res || (res[0] == '*')) {
+        throw std::system_error(errno, std::system_category());
     }
-    co_return res;
+    return res;
 }
 
 std::string_view prefix_for_scheme(scheme c) noexcept {
@@ -81,9 +58,8 @@ no_supported_schemes::no_supported_schemes()
         : std::runtime_error("No allowed hashing schemes are supported on this system") {
 }
 
-seastar::future<bool> check(const sstring& pass, const sstring& salted_hash) {
-    const auto pwd_hash = co_await detail::hash_with_salt_async(pass, salted_hash);
-    co_return pwd_hash == salted_hash;
+bool check(const sstring& pass, const sstring& salted_hash) {
+    return detail::hash_with_salt(pass, salted_hash) == salted_hash;
 }
 
 } // namespace auth::passwords

auth/passwords.hh

@@ -11,7 +11,6 @@
 #include <random>
 #include <stdexcept>
 
-#include <seastar/core/future.hh>
 #include <seastar/core/sstring.hh>
 
 #include "seastarx.hh"
@@ -76,19 +75,10 @@ sstring generate_salt(RandomNumberEngine& g, scheme scheme) {
 
 ///
 /// Hash a password combined with an implementation-specific salt string.
-/// Deprecated in favor of `hash_with_salt_async`.
 ///
 /// \throws \ref std::system_error when an unexpected implementation-specific error occurs.
 ///
-[[deprecated("Use hash_with_salt_async instead")]] sstring hash_with_salt(const sstring& pass, const sstring& salt);
-
-///
-/// Async version of `hash_with_salt` that returns a future.
-/// If possible, hashing uses `coroutine::maybe_yield` to prevent reactor stalls.
-///
-/// \throws \ref std::system_error when an unexpected implementation-specific error occurs.
-///
-seastar::future<sstring> hash_with_salt_async(const sstring& pass, const sstring& salt);
+sstring hash_with_salt(const sstring& pass, const sstring& salt);
 
 } // namespace detail
 
@@ -117,6 +107,6 @@ sstring hash(const sstring& pass, RandomNumberEngine& g, scheme scheme) {
 ///
 /// \throws \ref std::system_error when an unexpected implementation-specific error occurs.
 ///
-seastar::future<bool> check(const sstring& pass, const sstring& salted_hash);
+bool check(const sstring& pass, const sstring& salted_hash);
 
 } // namespace auth::passwords

auth/saslauthd_authenticator.cc

@@ -35,9 +35,10 @@ static const class_registrator<
         cql3::query_processor&,
         ::service::raft_group0_client&,
         ::service::migration_manager&,
-        cache&> saslauthd_auth_reg("com.scylladb.auth.SaslauthdAuthenticator");
+        cache&,
+        utils::alien_worker&> saslauthd_auth_reg("com.scylladb.auth.SaslauthdAuthenticator");
 
-saslauthd_authenticator::saslauthd_authenticator(cql3::query_processor& qp, ::service::raft_group0_client&, ::service::migration_manager&, cache&)
+saslauthd_authenticator::saslauthd_authenticator(cql3::query_processor& qp, ::service::raft_group0_client&, ::service::migration_manager&, cache&, utils::alien_worker&)
     : _socket_path(qp.db().get_config().saslauthd_socket_path())
 {}
 

auth/saslauthd_authenticator.hh

@@ -12,6 +12,7 @@
 
 #include "auth/authenticator.hh"
 #include "auth/cache.hh"
+#include "utils/alien_worker.hh"
 
 namespace cql3 {
 class query_processor;
@@ -29,7 +30,7 @@ namespace auth {
 class saslauthd_authenticator : public authenticator {
     sstring _socket_path; ///< Path to the domain socket on which saslauthd is listening.
 public:
-    saslauthd_authenticator(cql3::query_processor&, ::service::raft_group0_client&, ::service::migration_manager&, cache&);
+    saslauthd_authenticator(cql3::query_processor&, ::service::raft_group0_client&, ::service::migration_manager&, cache&,utils::alien_worker&);
 
     future<> start() override;
 

auth/service.cc

@@ -191,7 +191,8 @@ service::service(
         ::service::migration_manager& mm,
         const service_config& sc,
         maintenance_socket_enabled used_by_maintenance_socket,
-        cache& cache)
+        cache& cache,
+        utils::alien_worker& hashing_worker)
             : service(
                       std::move(c),
                       cache,
@@ -199,7 +200,7 @@ service::service(
                       g0,
                       mn,
                       create_object<authorizer>(sc.authorizer_java_name, qp, g0, mm),
-                      create_object<authenticator>(sc.authenticator_java_name, qp, g0, mm, cache),
+                      create_object<authenticator>(sc.authenticator_java_name, qp, g0, mm, cache, hashing_worker),
                       create_object<role_manager>(sc.role_manager_java_name, qp, g0, mm, cache),
                       used_by_maintenance_socket) {
 }
@@ -225,7 +226,7 @@ future<> service::create_legacy_keyspace_if_missing(::service::migration_manager
             try {
                 co_return co_await mm.announce(::service::prepare_new_keyspace_announcement(db.real_database(), ksm, ts),
                         std::move(group0_guard), seastar::format("auth_service: create {} keyspace", meta::legacy::AUTH_KS));
-            } catch (const ::service::group0_concurrent_modification&) {
+            } catch (::service::group0_concurrent_modification&) {
                 log.info("Concurrent operation is detected while creating {} keyspace, retrying.", meta::legacy::AUTH_KS);
             }
         }

auth/service.hh

@@ -27,6 +27,7 @@
 #include "cql3/description.hh"
 #include "seastarx.hh"
 #include "service/raft/raft_group0_client.hh"
+#include "utils/alien_worker.hh"
 #include "utils/observable.hh"
 #include "utils/serialized_action.hh"
 #include "service/maintenance_mode.hh"
@@ -130,7 +131,8 @@ public:
             ::service::migration_manager&,
             const service_config&,
             maintenance_socket_enabled,
-            cache&);
+            cache&,
+            utils::alien_worker&);
 
     future<> start(::service::migration_manager&, db::system_keyspace&);
 

auth/standard_role_manager.cc

@@ -192,7 +192,7 @@ future<> standard_role_manager::legacy_create_default_role_if_missing() {
                 {_superuser},
                 cql3::query_processor::cache_internal::no).discard_result();
         log.info("Created default superuser role '{}'.", _superuser);
-    } catch (const exceptions::unavailable_exception& e) {
+    } catch(const exceptions::unavailable_exception& e) {
         log.warn("Skipped default role setup: some nodes were not ready; will retry");
         throw e;
     }

auth/transitional.cc

@@ -38,8 +38,8 @@ class transitional_authenticator : public authenticator {
 public:
     static const sstring PASSWORD_AUTHENTICATOR_NAME;
 
-    transitional_authenticator(cql3::query_processor& qp, ::service::raft_group0_client& g0, ::service::migration_manager& mm, cache& cache)
-            : transitional_authenticator(std::make_unique<password_authenticator>(qp, g0, mm, cache)) {
+    transitional_authenticator(cql3::query_processor& qp, ::service::raft_group0_client& g0, ::service::migration_manager& mm, cache& cache, utils::alien_worker& hashing_worker)
+            : transitional_authenticator(std::make_unique<password_authenticator>(qp, g0, mm, cache, hashing_worker)) {
     }
     transitional_authenticator(std::unique_ptr<authenticator> a)
             : _authenticator(std::move(a)) {
@@ -81,7 +81,7 @@ public:
         }).handle_exception([](auto ep) {
             try {
                 std::rethrow_exception(ep);
-            } catch (const exceptions::authentication_exception&) {
+            } catch (exceptions::authentication_exception&) {
                 // return anon user
                 return make_ready_future<authenticated_user>(anonymous_user());
             }
@@ -126,7 +126,7 @@ public:
             virtual bytes evaluate_response(bytes_view client_response) override {
                 try {
                     return _sasl->evaluate_response(client_response);
-                } catch (const exceptions::authentication_exception&) {
+                } catch (exceptions::authentication_exception&) {
                     _complete = true;
                     return {};
                 }
@@ -141,7 +141,7 @@ public:
                     return _sasl->get_authenticated_user().handle_exception([](auto ep) {
                         try {
                             std::rethrow_exception(ep);
-                        } catch (const exceptions::authentication_exception&) {
+                        } catch (exceptions::authentication_exception&) {
                             // return anon user
                             return make_ready_future<authenticated_user>(anonymous_user());
                         }
@@ -241,7 +241,8 @@ static const class_registrator<
         cql3::query_processor&,
         ::service::raft_group0_client&,
         ::service::migration_manager&,
-        auth::cache&> transitional_authenticator_reg(auth::PACKAGE_NAME + "TransitionalAuthenticator");
+        auth::cache&,
+        utils::alien_worker&> transitional_authenticator_reg(auth::PACKAGE_NAME + "TransitionalAuthenticator");
 
 static const class_registrator<
         auth::authorizer,

client_data.hh

@@ -10,9 +10,7 @@
 #include <seastar/net/inet_address.hh>
 #include <seastar/core/sstring.hh>
 #include "seastarx.hh"
-#include "utils/loading_shared_values.hh"
 
-#include <list>
 #include <optional>
 
 enum class client_type {
@@ -29,20 +27,6 @@ enum class client_connection_stage {
     ready,
 };
 
-// We implement a keys cache using a map-like utils::loading_shared_values container by storing empty values.
-struct options_cache_value_type {};
-using client_options_cache_type = utils::loading_shared_values<sstring, options_cache_value_type>;
-using client_options_cache_entry_type = client_options_cache_type::entry_ptr;
-using client_options_cache_key_type = client_options_cache_type::key_type;
-
-// This struct represents a single OPTION key-value pair from the client's connection options.
-// Both key and value are represented by corresponding "references" to their cached values.
-// Each "reference" is effectively a lw_shared_ptr value.
-struct client_option_key_value_cached_entry {
-    client_options_cache_entry_type key;
-    client_options_cache_entry_type value;
-};
-
 sstring to_string(client_connection_stage ct);
 
 // Representation of a row in `system.clients'. std::optionals are for nullable cells.
@@ -53,8 +37,8 @@ struct client_data {
     client_connection_stage connection_stage = client_connection_stage::established;
     int32_t shard_id;  /// ID of server-side shard which is processing the connection.
 
-    std::optional<client_options_cache_entry_type> driver_name;
-    std::optional<client_options_cache_entry_type> driver_version;
+    std::optional<sstring> driver_name;
+    std::optional<sstring> driver_version;
     std::optional<sstring> hostname;
     std::optional<int32_t> protocol_version;
     std::optional<sstring> ssl_cipher_suite;
@@ -62,7 +46,6 @@ struct client_data {
     std::optional<sstring> ssl_protocol;
     std::optional<sstring> username;
     std::optional<sstring> scheduling_group_name;
-    std::list<client_option_key_value_cached_entry> client_options;
 
     sstring stage_str() const { return to_string(connection_stage); }
     sstring client_type_str() const { return to_string(ct); }

cmake/mode.common.cmake

@@ -125,6 +125,10 @@ if(target_arch)
   add_compile_options("-march=${target_arch}")
 endif()
 
+if(CMAKE_CXX_COMPILER_ID STREQUAL "Clang")
+  add_compile_options("SHELL:-Xclang -fexperimental-assignment-tracking=disabled")
+endif()
+
 function(maybe_limit_stack_usage_in_KB stack_usage_threshold_in_KB config)
   math(EXPR _stack_usage_threshold_in_bytes "${stack_usage_threshold_in_KB} * 1024")
   set(_stack_usage_threshold_flag "-Wstack-usage=${_stack_usage_threshold_in_bytes}")

configure.py

@@ -445,7 +445,6 @@ ldap_tests = set([
 scylla_tests = set([
     'test/boost/combined_tests',
     'test/boost/UUID_test',
-    'test/boost/url_parse_test',
     'test/boost/advanced_rpc_compressor_test',
     'test/boost/allocation_strategy_test',
     'test/boost/alternator_unit_test',
@@ -859,7 +858,6 @@ scylla_core = (['message/messaging_service.cc',
                 'utils/alien_worker.cc',
                 'utils/array-search.cc',
                 'utils/base64.cc',
-                'utils/crypt_sha512.cc',
                 'utils/logalloc.cc',
                 'utils/large_bitset.cc',
                 'utils/buffer_input_stream.cc',
@@ -1063,6 +1061,7 @@ scylla_core = (['message/messaging_service.cc',
                 'db/hints/resource_manager.cc',
                 'db/hints/sync_point.cc',
                 'db/large_data_handler.cc',
+                'db/legacy_schema_migrator.cc',
                 'db/marshal/type_parser.cc',
                 'db/per_partition_rate_limit_options.cc',
                 'db/rate_limiter.cc',
@@ -1158,7 +1157,6 @@ scylla_core = (['message/messaging_service.cc',
                 'locator/topology.cc',
                 'locator/util.cc',
                 'service/client_state.cc',
-                'service/client_routes.cc',
                 'service/storage_service.cc',
                 'service/session.cc',
                 'service/task_manager_module.cc',
@@ -1319,8 +1317,6 @@ api = ['api/api.cc',
        'api/storage_proxy.cc',
        Json2Code('api/api-doc/cache_service.json'),
        'api/cache_service.cc',
-       Json2Code('api/api-doc/client_routes.json'),
-       'api/client_routes.cc',
        Json2Code('api/api-doc/collectd.json'),
        'api/collectd.cc',
        Json2Code('api/api-doc/endpoint_snitch_info.json'),
@@ -1483,6 +1479,7 @@ deps = {
 
 pure_boost_tests = set([
     'test/boost/anchorless_list_test',
+    'test/boost/auth_passwords_test',
     'test/boost/auth_resource_test',
     'test/boost/big_decimal_test',
     'test/boost/caching_options_test',
@@ -1650,7 +1647,6 @@ deps['test/boost/bytes_ostream_test'] = [
 ]
 deps['test/boost/input_stream_test'] = ['test/boost/input_stream_test.cc']
 deps['test/boost/UUID_test'] = ['clocks-impl.cc', 'utils/UUID_gen.cc', 'test/boost/UUID_test.cc', 'utils/uuid.cc', 'utils/dynamic_bitset.cc', 'utils/hashers.cc', 'utils/on_internal_error.cc']
-deps['test/boost/url_parse_test'] = ['utils/http.cc', 'test/boost/url_parse_test.cc', ]
 deps['test/boost/murmur_hash_test'] = ['bytes.cc', 'utils/murmur_hash.cc', 'test/boost/murmur_hash_test.cc']
 deps['test/boost/allocation_strategy_test'] = ['test/boost/allocation_strategy_test.cc', 'utils/logalloc.cc', 'utils/dynamic_bitset.cc', 'utils/labels.cc']
 deps['test/boost/log_heap_test'] = ['test/boost/log_heap_test.cc']
@@ -2251,6 +2247,15 @@ def get_extra_cxxflags(mode, mode_config, cxx, debuginfo):
     if debuginfo and mode_config['can_have_debug_info']:
         cxxflags += ['-g', '-gz']
 
+    if 'clang' in cxx:
+        # Since AssignmentTracking was enabled by default in clang
+        # (llvm/llvm-project@de6da6ad55d3ca945195d1cb109cb8efdf40a52a)
+        # coroutine frame debugging info (`coro_frame_ty`) is broken.
+        #
+        # It seems that we aren't losing much by disabling AssigmentTracking,
+        # so for now we choose to disable it to get `coro_frame_ty` back.
+        cxxflags.append('-Xclang -fexperimental-assignment-tracking=disabled')
+
     return cxxflags
 
 

cql3/Cql.g

@@ -575,15 +575,6 @@ usingTimeoutServiceLevelClauseObjective[std::unique_ptr<cql3::attributes::raw>&
     | serviceLevel sl_name=serviceLevelOrRoleName { attrs->service_level = std::move(sl_name); }
     ;
 
-usingTimeoutConcurrencyClause[std::unique_ptr<cql3::attributes::raw>& attrs]
-    : K_USING usingTimeoutConcurrencyClauseObjective[attrs] ( K_AND usingTimeoutConcurrencyClauseObjective[attrs] )*
-    ;
-
-usingTimeoutConcurrencyClauseObjective[std::unique_ptr<cql3::attributes::raw>& attrs]
-    : K_TIMEOUT to=term { attrs->timeout = std::move(to); }
-    | K_CONCURRENCY c=term { attrs->concurrency = std::move(c); }
-    ;
-
 /**
  * UPDATE <CF>
  * USING TIMESTAMP <long>
@@ -675,7 +666,7 @@ pruneMaterializedViewStatement returns [std::unique_ptr<raw::select_statement> e
         auto attrs = std::make_unique<cql3::attributes::raw>();
         expression wclause = conjunction{};
     }
-	: K_PRUNE K_MATERIALIZED K_VIEW cf=columnFamilyName (K_WHERE w=whereClause { wclause = std::move(w); } )? ( usingTimeoutConcurrencyClause[attrs] )?
+	: K_PRUNE K_MATERIALIZED K_VIEW cf=columnFamilyName (K_WHERE w=whereClause { wclause = std::move(w); } )? ( usingClause[attrs] )?
 	  {
 	        auto params = make_lw_shared<raw::select_statement::parameters>(std::move(orderings), is_distinct, allow_filtering, statement_subtype, bypass_cache);
 	        return std::make_unique<raw::select_statement>(std::move(cf), std::move(params),
@@ -2379,7 +2370,6 @@ K_LIKE:        L I K E;
 
 K_TIMEOUT:     T I M E O U T;
 K_PRUNE:       P R U N E;
-K_CONCURRENCY: C O N C U R R E N C Y;
 
 K_EXECUTE:     E X E C U T E;
 

cql3/attributes.cc

@@ -20,21 +20,19 @@
 namespace cql3 {
 
 std::unique_ptr<attributes> attributes::none() {
-    return std::unique_ptr<attributes>{new attributes{{}, {}, {}, {}, {}}};
+    return std::unique_ptr<attributes>{new attributes{{}, {}, {}, {}}};
 }
 
 attributes::attributes(std::optional<cql3::expr::expression>&& timestamp,
                        std::optional<cql3::expr::expression>&& time_to_live,
                        std::optional<cql3::expr::expression>&& timeout,
-                       std::optional<sstring> service_level,
-                       std::optional<cql3::expr::expression>&& concurrency)
+                       std::optional<sstring> service_level)
     : _timestamp_unset_guard(timestamp)
     , _timestamp{std::move(timestamp)}
     , _time_to_live_unset_guard(time_to_live)
     , _time_to_live{std::move(time_to_live)}
     , _timeout{std::move(timeout)}
     , _service_level(std::move(service_level))
-    , _concurrency{std::move(concurrency)}
 { }
 
 bool attributes::is_timestamp_set() const {
@@ -53,10 +51,6 @@ bool attributes::is_service_level_set() const {
     return bool(_service_level);
 }
 
-bool attributes::is_concurrency_set() const {
-    return bool(_concurrency);
-}
-
 int64_t attributes::get_timestamp(int64_t now, const query_options& options) {
     if (!_timestamp.has_value() || _timestamp_unset_guard.is_unset(options)) {
         return now;
@@ -129,27 +123,6 @@ qos::service_level_options attributes::get_service_level(qos::service_level_cont
     return sl_controller.get_service_level(sl_name).slo;
 }
 
-std::optional<int32_t> attributes::get_concurrency(const query_options& options) const {
-    if (!_concurrency.has_value()) {
-        return std::nullopt;
-    }
-
-    cql3::raw_value concurrency_raw = expr::evaluate(*_concurrency, options);
-    if (concurrency_raw.is_null()) {
-        throw exceptions::invalid_request_exception("Invalid null value of concurrency");
-    }
-    int32_t concurrency;
-    try {
-        concurrency = concurrency_raw.view().validate_and_deserialize<int32_t>(*int32_type);
-    } catch (marshal_exception& e) {
-        throw exceptions::invalid_request_exception("Invalid concurrency value");
-    }
-    if (concurrency <= 0) {
-        throw exceptions::invalid_request_exception("Concurrency must be a positive integer");
-    }
-    return concurrency;
-}
-
 void attributes::fill_prepare_context(prepare_context& ctx) {
     if (_timestamp.has_value()) {
         expr::fill_prepare_context(*_timestamp, ctx);
@@ -160,13 +133,10 @@ void attributes::fill_prepare_context(prepare_context& ctx) {
     if (_timeout.has_value()) {
         expr::fill_prepare_context(*_timeout, ctx);
     }
-    if (_concurrency.has_value()) {
-        expr::fill_prepare_context(*_concurrency, ctx);
-    }
 }
 
 std::unique_ptr<attributes> attributes::raw::prepare(data_dictionary::database db, const sstring& ks_name, const sstring& cf_name) const {
-    std::optional<expr::expression> ts, ttl, to, conc;
+    std::optional<expr::expression> ts, ttl, to;
 
     if (timestamp.has_value()) {
         ts = prepare_expression(*timestamp, db, ks_name, nullptr, timestamp_receiver(ks_name, cf_name));
@@ -183,12 +153,7 @@ std::unique_ptr<attributes> attributes::raw::prepare(data_dictionary::database d
         verify_no_aggregate_functions(*timeout, "USING clause");
     }
 
-    if (concurrency.has_value()) {
-        conc = prepare_expression(*concurrency, db, ks_name, nullptr, concurrency_receiver(ks_name, cf_name));
-        verify_no_aggregate_functions(*concurrency, "USING clause");
-    }
-
-    return std::unique_ptr<attributes>{new attributes{std::move(ts), std::move(ttl), std::move(to), std::move(service_level), std::move(conc)}};
+    return std::unique_ptr<attributes>{new attributes{std::move(ts), std::move(ttl), std::move(to), std::move(service_level)}};
 }
 
 lw_shared_ptr<column_specification> attributes::raw::timestamp_receiver(const sstring& ks_name, const sstring& cf_name) const {
@@ -203,8 +168,4 @@ lw_shared_ptr<column_specification> attributes::raw::timeout_receiver(const sstr
     return make_lw_shared<column_specification>(ks_name, cf_name, ::make_shared<column_identifier>("[timeout]", true), duration_type);
 }
 
-lw_shared_ptr<column_specification> attributes::raw::concurrency_receiver(const sstring& ks_name, const sstring& cf_name) const {
-    return make_lw_shared<column_specification>(ks_name, cf_name, ::make_shared<column_identifier>("[concurrency]", true), data_type_for<int32_t>());
-}
-
 }

cql3/attributes.hh

@@ -36,15 +36,13 @@ private:
     std::optional<cql3::expr::expression> _time_to_live;
     std::optional<cql3::expr::expression> _timeout;
     std::optional<sstring> _service_level;
-    std::optional<cql3::expr::expression> _concurrency;
 public:
     static std::unique_ptr<attributes> none();
 private:
     attributes(std::optional<cql3::expr::expression>&& timestamp,
                std::optional<cql3::expr::expression>&& time_to_live,
                std::optional<cql3::expr::expression>&& timeout,
-               std::optional<sstring> service_level,
-               std::optional<cql3::expr::expression>&& concurrency);
+               std::optional<sstring> service_level);
 public:
     bool is_timestamp_set() const;
 
@@ -54,8 +52,6 @@ public:
 
     bool is_service_level_set() const;
 
-    bool is_concurrency_set() const;
-
     int64_t get_timestamp(int64_t now, const query_options& options);
 
     std::optional<int32_t> get_time_to_live(const query_options& options);
@@ -64,8 +60,6 @@ public:
 
     qos::service_level_options get_service_level(qos::service_level_controller& sl_controller) const;
 
-    std::optional<int32_t> get_concurrency(const query_options& options) const;
-
     void fill_prepare_context(prepare_context& ctx);
 
     class raw final {
@@ -74,7 +68,6 @@ public:
         std::optional<cql3::expr::expression> time_to_live;
         std::optional<cql3::expr::expression> timeout;
         std::optional<sstring> service_level;
-        std::optional<cql3::expr::expression> concurrency;
 
         std::unique_ptr<attributes> prepare(data_dictionary::database db, const sstring& ks_name, const sstring& cf_name) const;
     private:
@@ -83,8 +76,6 @@ public:
         lw_shared_ptr<column_specification> time_to_live_receiver(const sstring& ks_name, const sstring& cf_name) const;
 
         lw_shared_ptr<column_specification> timeout_receiver(const sstring& ks_name, const sstring& cf_name) const;
-
-        lw_shared_ptr<column_specification> concurrency_receiver(const sstring& ks_name, const sstring& cf_name) const;
     };
 };
 

cql3/query_processor.cc

@@ -64,10 +64,6 @@ bool query_processor::topology_global_queue_empty() {
     return remote().first.get().ss.topology_global_queue_empty();
 }
 
-future<bool> query_processor::ongoing_rf_change(const service::group0_guard& guard, sstring ks) {
-    return remote().first.get().ss.ongoing_rf_change(guard, std::move(ks));
-}
-
 static service::query_state query_state_for_internal_call() {
     return {service::client_state::for_internal_calls(), empty_service_permit()};
 }

cql3/query_processor.hh

@@ -474,7 +474,6 @@ public:
     void reset_cache();
 
     bool topology_global_queue_empty();
-    future<bool> ongoing_rf_change(const service::group0_guard& guard, sstring ks);
 
     query_options make_internal_options(
             const statements::prepared_statement::checked_weak_ptr& p,

cql3/restrictions/statement_restrictions.cc

@@ -1322,10 +1322,6 @@ const std::vector<expr::expression>& statement_restrictions::index_restrictions(
     return _index_restrictions;
 }
 
-bool statement_restrictions::is_empty() const {
-    return !_where.has_value();
-}
-
 // Current score table:
 // local and restrictions include full partition key: 2
 // global: 1

cql3/restrictions/statement_restrictions.hh

@@ -408,8 +408,6 @@ public:
 
     /// Checks that the primary key restrictions don't contain null values, throws invalid_request_exception otherwise.
     void validate_primary_key(const query_options& options) const;
-
-    bool is_empty() const;
 };
 
 statement_restrictions analyze_statement_restrictions(

cql3/statements/alter_keyspace_statement.cc

@@ -19,7 +19,6 @@
 #include "locator/abstract_replication_strategy.hh"
 #include "mutation/canonical_mutation.hh"
 #include "prepared_statement.hh"
-#include "seastar/coroutine/exception.hh"
 #include "service/migration_manager.hh"
 #include "service/storage_proxy.hh"
 #include "service/topology_mutation.hh"
@@ -139,7 +138,6 @@ bool cql3::statements::alter_keyspace_statement::changes_tablets(query_processor
 future<std::tuple<::shared_ptr<cql_transport::event::schema_change>, cql3::cql_warnings_vec>>
 cql3::statements::alter_keyspace_statement::prepare_schema_mutations(query_processor& qp, service::query_state& state, const query_options& options, service::group0_batch& mc) const {
     using namespace cql_transport;
-    bool unknown_keyspace = false;
     try {
         event::schema_change::target_type target_type = event::schema_change::target_type::KEYSPACE;
         auto ks = qp.db().find_keyspace(_name);
@@ -160,12 +158,8 @@ cql3::statements::alter_keyspace_statement::prepare_schema_mutations(query_proce
         //       when in reality nothing or only schema is being changed
         if (changes_tablets(qp)) {
             if (!qp.proxy().features().topology_global_request_queue && !qp.topology_global_queue_empty()) {
-                co_await coroutine::return_exception(
-                    exceptions::invalid_request_exception("Another global topology request is ongoing, please retry."));
-            }
-            if (qp.proxy().features().rack_list_rf && co_await qp.ongoing_rf_change(mc.guard(),_name)) {
-                co_await coroutine::return_exception(
-                        exceptions::invalid_request_exception(format("Another RF change for this keyspace {} ongoing, please retry.", _name)));
+                return make_exception_future<std::tuple<::shared_ptr<::cql_transport::event::schema_change>, cql3::cql_warnings_vec>>(
+                        exceptions::invalid_request_exception("Another global topology request is ongoing, please retry."));
             }
             qp.db().real_database().validate_keyspace_update(*ks_md_update);
 
@@ -248,15 +242,10 @@ cql3::statements::alter_keyspace_statement::prepare_schema_mutations(query_proce
                 target_type,
                 keyspace());
         mc.add_mutations(std::move(muts), "CQL alter keyspace");
-        co_return std::make_tuple(std::move(ret), warnings);
+        return make_ready_future<std::tuple<::shared_ptr<cql_transport::event::schema_change>, cql3::cql_warnings_vec>>(std::make_tuple(std::move(ret), warnings));
     } catch (data_dictionary::no_such_keyspace& e) {
-        unknown_keyspace = true;
+        return make_exception_future<std::tuple<::shared_ptr<cql_transport::event::schema_change>, cql3::cql_warnings_vec>>(exceptions::invalid_request_exception("Unknown keyspace " + _name));
     }
-    if (unknown_keyspace) {
-        co_await coroutine::return_exception(
-                exceptions::invalid_request_exception("Unknown keyspace " + _name));
-    }
-    std::unreachable();
 }
 
 std::unique_ptr<cql3::statements::prepared_statement>

cql3/statements/batch_statement.cc

@@ -331,7 +331,7 @@ future<shared_ptr<cql_transport::messages::result_message>> batch_statement::exe
     if (!cl_for_paxos) [[unlikely]] {
         return make_exception_future<shared_ptr<cql_transport::messages::result_message>>(std::move(cl_for_paxos).assume_error());
     }
-    std::unique_ptr<cas_request> request;
+    seastar::shared_ptr<cas_request> request;
     schema_ptr schema;
 
     db::timeout_clock::time_point now = db::timeout_clock::now();
@@ -354,9 +354,9 @@ future<shared_ptr<cql_transport::messages::result_message>> batch_statement::exe
         if (keys.empty()) {
             continue;
         }
-        if (!request) {
+        if (request.get() == nullptr) {
             schema = statement.s;
-            request = std::make_unique<cas_request>(schema, std::move(keys));
+            request = seastar::make_shared<cas_request>(schema, std::move(keys));
         } else if (keys.size() != 1 || keys.front().equal(request->key().front(), dht::ring_position_comparator(*schema)) == false) {
             throw exceptions::invalid_request_exception("BATCH with conditions cannot span multiple partitions");
         }
@@ -366,7 +366,7 @@ future<shared_ptr<cql_transport::messages::result_message>> batch_statement::exe
 
         request->add_row_update(statement, std::move(ranges), std::move(json_cache), statement_options);
     }
-    if (!request) {
+    if (request.get() == nullptr) {
         throw exceptions::invalid_request_exception(format("Unrestricted partition key in a conditional BATCH"));
     }
 
@@ -377,10 +377,9 @@ future<shared_ptr<cql_transport::messages::result_message>> batch_statement::exe
             );
     }
 
-    auto* request_ptr = request.get();
-    return qp.proxy().cas(schema, std::move(cas_shard), *request_ptr, request->read_command(qp), request->key(),
+    return qp.proxy().cas(schema, std::move(cas_shard), request, request->read_command(qp), request->key(),
             {read_timeout, qs.get_permit(), qs.get_client_state(), qs.get_trace_state()},
-            std::move(cl_for_paxos).assume_value(), cl_for_learn, batch_timeout, cas_timeout).then([this, request = std::move(request)] (bool is_applied) {
+            std::move(cl_for_paxos).assume_value(), cl_for_learn, batch_timeout, cas_timeout).then([this, request] (bool is_applied) {
         return request->build_cas_result_set(_metadata, _columns_of_cas_result_set, is_applied);
     });
 }

cql3/statements/create_index_statement.cc

@@ -279,15 +279,11 @@ std::vector<::shared_ptr<index_target>> create_index_statement::validate_while_e
         throw exceptions::invalid_request_exception(format("index names shouldn't be more than {:d} characters long (got \"{}\")", schema::NAME_LENGTH, _index_name.c_str()));
     }
 
-    // Regular secondary indexes require rf-rack-validity.
-    // Custom indexes need to validate this property themselves, if they need it.
-    if (!_properties || !_properties->custom_class) {
-        try {
-            db::view::validate_view_keyspace(db, keyspace());
-        } catch (const std::exception& e) {
-            // The type of the thrown exception is not specified, so we need to wrap it here.
-            throw exceptions::invalid_request_exception(e.what());
-        }
+    try {
+        db::view::validate_view_keyspace(db, keyspace());
+    } catch (const std::exception& e) {
+        // The type of the thrown exception is not specified, so we need to wrap it here.
+        throw exceptions::invalid_request_exception(e.what());
     }
 
     validate_for_local_index(*schema);

cql3/statements/ks_prop_defs.cc

@@ -61,7 +61,7 @@ expand_to_racks(const locator::token_metadata& tm,
 
     // Handle ALTER:
     // ([]|0) -> numeric is allowed, there are no existing replicas
-    // numeric -> numeric' is not supported unless numeric == numeric'. User should convert RF to rack list of equal count first.
+    // numeric -> numeric' is not supported. User should convert RF to rack list of equal count first.
     // rack_list -> len(rack_list) is allowed (no-op)
     // rack_list -> numeric is not allowed
     if (old_options.contains(dc)) {
@@ -75,8 +75,6 @@ expand_to_racks(const locator::token_metadata& tm,
                         "Cannot change replication factor for '{}' from {} to numeric {}, use rack list instead",
                         dc, old_rf_val, data.count()));
             }
-        } else if (old_rf.count() == data.count()) {
-            return rf;
         } else if (old_rf.count() > 0) {
             throw exceptions::configuration_exception(fmt::format(
                     "Cannot change replication factor for '{}' from {} to {}, only rack list is allowed",
@@ -155,8 +153,6 @@ static locator::replication_strategy_config_options prepare_options(
     }
 
     // Validate options.
-    bool numeric_to_rack_list_transition = false;
-    bool rf_change = false;
     for (auto&& [dc, opt] : options) {
         locator::replication_factor_data rf(opt);
 
@@ -166,7 +162,6 @@ static locator::replication_strategy_config_options prepare_options(
             old_rf = locator::replication_factor_data(i->second);
         }
 
-        rf_change = rf_change || (old_rf && old_rf->count() != rf.count()) || (!old_rf && rf.count() != 0);
         if (!rf.is_rack_based()) {
             if (old_rf && old_rf->is_rack_based() && rf.count() != 0) {
                 if (old_rf->count() != rf.count()) {
@@ -192,11 +187,12 @@ static locator::replication_strategy_config_options prepare_options(
             throw exceptions::configuration_exception(fmt::format(
                     "Rack list for '{}' contains duplicate entries", dc));
         }
-        numeric_to_rack_list_transition = numeric_to_rack_list_transition || (old_rf && !old_rf->is_rack_based() && old_rf->count() != 0);
-    }
-
-    if (numeric_to_rack_list_transition && rf_change) {
-        throw exceptions::configuration_exception("Cannot change replication factor from numeric to rack list and rf value at the same time");
+        if (old_rf && !old_rf->is_rack_based() && old_rf->count() != 0) {
+            // FIXME: Allow this if replicas already conform to the given rack list.
+            // FIXME: Implement automatic colocation to allow transition to rack list.
+            throw exceptions::configuration_exception(fmt::format(
+                    "Cannot change replication factor from numeric to rack list for '{}'", dc));
+        }
     }
 
     if (!rf && options.empty() && old_options.empty()) {
@@ -416,7 +412,7 @@ lw_shared_ptr<data_dictionary::keyspace_metadata> ks_prop_defs::as_ks_metadata(s
             ? std::optional<unsigned>(0) : std::nullopt;
     auto initial_tablets = get_initial_tablets(default_initial_tablets, cfg.enforce_tablets());
     bool uses_tablets = initial_tablets.has_value();
-    bool rack_list_enabled = utils::get_local_injector().enter("create_with_numeric") ? false : feat.rack_list_rf;
+    bool rack_list_enabled = feat.rack_list_rf;
     auto options = prepare_options(sc, tm, cfg.rf_rack_valid_keyspaces(), get_replication_options(), {}, rack_list_enabled, uses_tablets);
     return data_dictionary::keyspace_metadata::new_keyspace(ks_name, sc,
             std::move(options), initial_tablets, get_consistency_option(), get_boolean(KW_DURABLE_WRITES, true), get_storage_options());
@@ -432,7 +428,7 @@ lw_shared_ptr<data_dictionary::keyspace_metadata> ks_prop_defs::as_ks_metadata_u
         throw exceptions::invalid_request_exception("Cannot alter replication strategy vnode/tablets flavor");
     }
     auto sc = get_replication_strategy_class();
-    bool rack_list_enabled = utils::get_local_injector().enter("create_with_numeric") ? false : feat.rack_list_rf;
+    bool rack_list_enabled = feat.rack_list_rf;
     if (sc) {
         options = prepare_options(*sc, tm, cfg.rf_rack_valid_keyspaces(), get_replication_options(), old_options, rack_list_enabled, uses_tablets);
     } else {

cql3/statements/modification_statement.cc

@@ -401,8 +401,7 @@ modification_statement::execute_with_condition(query_processor& qp, service::que
                     type.is_update() ? "update" : "deletion"));
     }
 
-    auto request = std::make_unique<cas_request>(s, std::move(keys));
-    auto* request_ptr = request.get();
+    auto request = seastar::make_shared<cas_request>(s, std::move(keys));
     // cas_request can be used for batches as well single statements; Here we have just a single
     // modification in the list of CAS commands, since we're handling single-statement execution.
     request->add_row_update(*this, std::move(ranges), std::move(json_cache), options);
@@ -428,9 +427,9 @@ modification_statement::execute_with_condition(query_processor& qp, service::que
         tablet_info = erm->check_locality(token);
     }
 
-    return qp.proxy().cas(s, std::move(cas_shard), *request_ptr, request->read_command(qp), request->key(),
+    return qp.proxy().cas(s, std::move(cas_shard), request, request->read_command(qp), request->key(),
             {read_timeout, qs.get_permit(), qs.get_client_state(), qs.get_trace_state()},
-            std::move(cl_for_paxos).assume_value(), cl_for_learn, statement_timeout, cas_timeout).then([this, request = std::move(request), tablet_replicas = std::move(tablet_info->tablet_replicas), token_range = tablet_info->token_range] (bool is_applied) {
+            std::move(cl_for_paxos).assume_value(), cl_for_learn, statement_timeout, cas_timeout).then([this, request, tablet_replicas = std::move(tablet_info->tablet_replicas), token_range = tablet_info->token_range] (bool is_applied) {
         auto result = request->build_cas_result_set(_metadata, _columns_of_cas_result_set, is_applied);
         result->add_tablet_info(tablet_replicas, token_range);
         return result;

cql3/statements/prune_materialized_view_statement.cc

@@ -21,7 +21,7 @@ namespace cql3 {
 namespace statements {
 
 static future<> delete_ghost_rows(dht::partition_range_vector partition_ranges, std::vector<query::clustering_range> clustering_bounds, view_ptr view,
-        service::storage_proxy& proxy, service::query_state& state, const query_options& options, cql_stats& stats, db::timeout_clock::duration timeout_duration, size_t concurrency) {
+        service::storage_proxy& proxy, service::query_state& state, const query_options& options, cql_stats& stats, db::timeout_clock::duration timeout_duration) {
     auto key_columns = std::ranges::to<std::vector<const column_definition*>>(
         view->all_columns()
         | std::views::filter([] (const column_definition& cdef) { return cdef.is_primary_key(); })
@@ -35,7 +35,7 @@ static future<> delete_ghost_rows(dht::partition_range_vector partition_ranges,
     tracing::trace(state.get_trace_state(), "Deleting ghost rows from partition ranges {}", partition_ranges);
 
     auto p = service::pager::query_pagers::ghost_row_deleting_pager(schema_ptr(view), selection, state,
-            options, std::move(command), std::move(partition_ranges), stats, proxy, timeout_duration, concurrency);
+            options, std::move(command), std::move(partition_ranges), stats, proxy, timeout_duration);
 
     int32_t page_size = std::max(options.get_page_size(), 1000);
     auto now = gc_clock::now();
@@ -62,8 +62,7 @@ future<::shared_ptr<cql_transport::messages::result_message>> prune_materialized
     auto timeout_duration = get_timeout(state.get_client_state(), options);
     dht::partition_range_vector key_ranges = _restrictions->get_partition_key_ranges(options);
     std::vector<query::clustering_range> clustering_bounds = _restrictions->get_clustering_bounds(options);
-    size_t concurrency = _attrs->is_concurrency_set() ? _attrs->get_concurrency(options).value() : 1;
-    return delete_ghost_rows(std::move(key_ranges), std::move(clustering_bounds), view_ptr(_schema), qp.proxy(), state, options, _stats, timeout_duration, concurrency).then([] {
+    return delete_ghost_rows(std::move(key_ranges), std::move(clustering_bounds), view_ptr(_schema), qp.proxy(), state, options, _stats, timeout_duration).then([] {
         return make_ready_future<::shared_ptr<cql_transport::messages::result_message>>(::make_shared<cql_transport::messages::result_message::void_message>());
     });
 }

cql3/statements/select_statement.cc

@@ -1976,7 +1976,7 @@ mutation_fragments_select_statement::do_execute(query_processor& qp, service::qu
     if (it == indexes.end()) {
         throw exceptions::invalid_request_exception("ANN ordering by vector requires the column to be indexed using 'vector_index'");
     }
-    if (index_opt || parameters->allow_filtering() || !(restrictions->is_empty()) || check_needs_allow_filtering_anyway(*restrictions)) {
+    if (index_opt || parameters->allow_filtering() || restrictions->need_filtering() || check_needs_allow_filtering_anyway(*restrictions)) {
         throw exceptions::invalid_request_exception("ANN ordering by vector does not support filtering");
     }
     index_opt = *it;

data_dictionary/data_dictionary.cc

@@ -42,11 +42,6 @@ table::get_index_manager() const {
     return _ops->get_index_manager(*this);
 }
 
-db_clock::time_point
-table::get_truncation_time() const {
-    return _ops->get_truncation_time(*this);
-}
-
 lw_shared_ptr<keyspace_metadata>
 keyspace::metadata() const {
     return _ops->get_keyspace_metadata(*this);

data_dictionary/data_dictionary.hh

@@ -77,7 +77,6 @@ public:
     schema_ptr schema() const;
     const std::vector<view_ptr>& views() const;
     const secondary_index::secondary_index_manager& get_index_manager() const;
-    db_clock::time_point get_truncation_time() const;
 };
 
 class keyspace {

data_dictionary/impl.hh

@@ -27,7 +27,6 @@ public:
     virtual std::optional<table> try_find_table(database db, table_id id) const = 0;
     virtual const secondary_index::secondary_index_manager& get_index_manager(table t) const = 0;
     virtual schema_ptr get_table_schema(table t) const = 0;
-    virtual db_clock::time_point get_truncation_time(table t) const = 0;
     virtual lw_shared_ptr<keyspace_metadata> get_keyspace_metadata(keyspace ks) const = 0;
     virtual bool is_internal(keyspace ks) const = 0;
     virtual const locator::abstract_replication_strategy& get_replication_strategy(keyspace ks) const = 0;

db/CMakeLists.txt

@@ -10,6 +10,7 @@ target_sources(db
     schema_applier.cc
     schema_tables.cc
     cql_type_parser.cc
+    legacy_schema_migrator.cc
     commitlog/commitlog.cc
     commitlog/commitlog_replayer.cc
     commitlog/commitlog_entry.cc

db/batchlog.hh

@@ -1,20 +0,0 @@
-/*
- * Copyright (C) 2025-present ScyllaDB
- */
-
-/*
- * SPDX-License-Identifier: (LicenseRef-ScyllaDB-Source-Available-1.0 and Apache-2.0)
- */
-
-#pragma once
-
-#include "mutation/mutation.hh"
-#include "utils/UUID.hh"
-
-namespace db {
-
-mutation get_batchlog_mutation_for(schema_ptr schema, const utils::chunked_vector<mutation>& mutations, int32_t version, db_clock::time_point now, const utils::UUID& id);
-
-mutation get_batchlog_delete_mutation(schema_ptr schema, int32_t version, db_clock::time_point now, const utils::UUID& id);
-
-}

db/batchlog_manager.cc

@@ -10,7 +10,6 @@
 
 #include <chrono>
 #include <exception>
-#include <ranges>
 #include <seastar/core/future-util.hh>
 #include <seastar/core/do_with.hh>
 #include <seastar/core/semaphore.hh>
@@ -19,14 +18,12 @@
 #include <seastar/core/sleep.hh>
 
 #include "batchlog_manager.hh"
-#include "batchlog.hh"
 #include "data_dictionary/data_dictionary.hh"
 #include "mutation/canonical_mutation.hh"
 #include "service/storage_proxy.hh"
 #include "system_keyspace.hh"
 #include "utils/rate_limiter.hh"
 #include "utils/log.hh"
-#include "utils/murmur_hash.hh"
 #include "db_clock.hh"
 #include "unimplemented.hh"
 #include "idl/frozen_schema.dist.hh"
@@ -36,94 +33,17 @@
 #include "cql3/untyped_result_set.hh"
 #include "service_permit.hh"
 #include "cql3/query_processor.hh"
+#include "replica/database.hh"
 
 static logging::logger blogger("batchlog_manager");
 
-namespace db {
-
-// Yields 256 batchlog shards. Even on the largest nodes we currently run on,
-// this should be enough to give every core a batchlog partition.
-static constexpr unsigned batchlog_shard_bits = 8;
-
-int32_t batchlog_shard_of(db_clock::time_point written_at) {
-    const int64_t count = written_at.time_since_epoch().count();
-    std::array<uint64_t, 2> result;
-    utils::murmur_hash::hash3_x64_128(bytes_view(reinterpret_cast<const signed char*>(&count), sizeof(count)), 0, result);
-    uint64_t hash = result[0] ^ result[1];
-    return hash & ((1ULL << batchlog_shard_bits) - 1);
-}
-
-std::pair<partition_key, clustering_key>
-get_batchlog_key(const schema& schema, int32_t version, db::batchlog_stage stage, int32_t batchlog_shard, db_clock::time_point written_at, std::optional<utils::UUID> id) {
-    auto pkey = partition_key::from_exploded(schema, {serialized(version), serialized(int8_t(stage)), serialized(batchlog_shard)});
-
-    std::vector<bytes> ckey_components;
-    ckey_components.reserve(2);
-    ckey_components.push_back(serialized(written_at));
-    if (id) {
-        ckey_components.push_back(serialized(*id));
-    }
-    auto ckey = clustering_key::from_exploded(schema, ckey_components);
-
-    return {std::move(pkey), std::move(ckey)};
-}
-
-std::pair<partition_key, clustering_key>
-get_batchlog_key(const schema& schema, int32_t version, db::batchlog_stage stage, db_clock::time_point written_at, std::optional<utils::UUID> id) {
-    return get_batchlog_key(schema, version, stage, batchlog_shard_of(written_at), written_at, id);
-}
-
-mutation get_batchlog_mutation_for(schema_ptr schema, managed_bytes data, int32_t version, db::batchlog_stage stage, db_clock::time_point now, const utils::UUID& id) {
-    auto [key, ckey] = get_batchlog_key(*schema, version, stage, now, id);
-
-    auto timestamp = api::new_timestamp();
-
-    mutation m(schema, key);
-    // Avoid going through data_value and therefore `bytes`, as it can be large (#24809).
-    auto cdef_data = schema->get_column_definition(to_bytes("data"));
-    m.set_cell(ckey, *cdef_data, atomic_cell::make_live(*cdef_data->type, timestamp, std::move(data)));
-
-    return m;
-}
-
-mutation get_batchlog_mutation_for(schema_ptr schema, const utils::chunked_vector<mutation>& mutations, int32_t version, db::batchlog_stage stage, db_clock::time_point now, const utils::UUID& id) {
-    auto data = [&mutations] {
-        utils::chunked_vector<canonical_mutation> fm(mutations.begin(), mutations.end());
-        bytes_ostream out;
-        for (auto& m : fm) {
-            ser::serialize(out, m);
-        }
-        return std::move(out).to_managed_bytes();
-    }();
-
-    return get_batchlog_mutation_for(std::move(schema), std::move(data), version, stage, now, id);
-}
-
-mutation get_batchlog_mutation_for(schema_ptr schema, const utils::chunked_vector<mutation>& mutations, int32_t version, db_clock::time_point now, const utils::UUID& id) {
-    return get_batchlog_mutation_for(std::move(schema), mutations, version, batchlog_stage::initial, now, id);
-}
-
-mutation get_batchlog_delete_mutation(schema_ptr schema, int32_t version, db::batchlog_stage stage, db_clock::time_point now, const utils::UUID& id) {
-    auto [key, ckey] = get_batchlog_key(*schema, version, stage, now, id);
-    mutation m(schema, key);
-    auto timestamp = api::new_timestamp();
-    m.partition().apply_delete(*schema, ckey, tombstone(timestamp, gc_clock::now()));
-    return m;
-}
-
-mutation get_batchlog_delete_mutation(schema_ptr schema, int32_t version, db_clock::time_point now, const utils::UUID& id) {
-    return get_batchlog_delete_mutation(std::move(schema), version, batchlog_stage::initial, now, id);
-}
-
-} // namespace db
-
 const std::chrono::seconds db::batchlog_manager::replay_interval;
 const uint32_t db::batchlog_manager::page_size;
 
 db::batchlog_manager::batchlog_manager(cql3::query_processor& qp, db::system_keyspace& sys_ks, batchlog_manager_config config)
         : _qp(qp)
         , _sys_ks(sys_ks)
-        , _replay_timeout(config.replay_timeout)
+        , _write_request_timeout(std::chrono::duration_cast<db_clock::duration>(config.write_request_timeout))
         , _replay_rate(config.replay_rate)
         , _delay(config.delay)
         , _replay_cleanup_after_replays(config.replay_cleanup_after_replays)
@@ -232,75 +152,18 @@ future<> db::batchlog_manager::stop() {
 }
 
 future<size_t> db::batchlog_manager::count_all_batches() const {
-    sstring query = format("SELECT count(*) FROM {}.{} BYPASS CACHE", system_keyspace::NAME, system_keyspace::BATCHLOG_V2);
+    sstring query = format("SELECT count(*) FROM {}.{} BYPASS CACHE", system_keyspace::NAME, system_keyspace::BATCHLOG);
     return _qp.execute_internal(query, cql3::query_processor::cache_internal::yes).then([](::shared_ptr<cql3::untyped_result_set> rs) {
        return size_t(rs->one().get_as<int64_t>("count"));
     });
 }
 
-future<> db::batchlog_manager::maybe_migrate_v1_to_v2() {
-    if (_migration_done) {
-        return make_ready_future<>();
-    }
-    return with_gate(_gate, [this] () mutable -> future<> {
-        blogger.info("Migrating batchlog entries from v1 -> v2");
-
-        auto schema_v1 = _qp.db().find_schema(system_keyspace::NAME, system_keyspace::BATCHLOG);
-        auto schema_v2 = _qp.db().find_schema(system_keyspace::NAME, system_keyspace::BATCHLOG_V2);
-
-        auto batch = [this, schema_v1, schema_v2] (const cql3::untyped_result_set::row& row) -> future<stop_iteration> {
-            // check version of serialization format
-            if (!row.has("version")) {
-                blogger.warn("Not migrating logged batch because of unknown version");
-                co_return stop_iteration::no;
-            }
-
-            auto version = row.get_as<int32_t>("version");
-            if (version != netw::messaging_service::current_version) {
-                blogger.warn("Not migrating logged batch because of incorrect version");
-                co_return stop_iteration::no;
-            }
-
-            auto id = row.get_as<utils::UUID>("id");
-            auto written_at = row.get_as<db_clock::time_point>("written_at");
-            auto data = row.get_blob_fragmented("data");
-
-            auto& sp = _qp.proxy();
-
-            utils::get_local_injector().inject("batchlog_manager_fail_migration", [] { throw std::runtime_error("Error injection: failing batchlog migration"); });
-
-            auto migrate_mut = get_batchlog_mutation_for(schema_v2, std::move(data), version, batchlog_stage::failed_replay, written_at, id);
-            co_await sp.mutate_locally(migrate_mut, tracing::trace_state_ptr(), db::commitlog::force_sync::no);
-
-            mutation delete_mut(schema_v1, partition_key::from_single_value(*schema_v1, serialized(id)));
-            delete_mut.partition().apply_delete(*schema_v1, clustering_key_prefix::make_empty(), tombstone(api::new_timestamp(), gc_clock::now()));
-            co_await sp.mutate_locally(delete_mut, tracing::trace_state_ptr(), db::commitlog::force_sync::no);
-
-            co_return stop_iteration::no;
-        };
-        try {
-            co_await _qp.query_internal(
-                    format("SELECT * FROM {}.{} BYPASS CACHE", system_keyspace::NAME, system_keyspace::BATCHLOG),
-                    db::consistency_level::ONE,
-                    {},
-                    page_size,
-                    std::move(batch));
-        } catch (...) {
-            blogger.warn("Batchlog v1 to v2 migration failed: {}; will retry", std::current_exception());
-            co_return;
-        }
-
-        co_await container().invoke_on_all([] (auto& bm) {
-            bm._migration_done = true;
-        });
-
-        blogger.info("Done migrating batchlog entries from v1 -> v2");
-    });
+db_clock::duration db::batchlog_manager::get_batch_log_timeout() const {
+    // enough time for the actual write + BM removal mutation
+    return _write_request_timeout * 2;
 }
 
 future<db::all_batches_replayed> db::batchlog_manager::replay_all_failed_batches(post_replay_cleanup cleanup) {
-    co_await maybe_migrate_v1_to_v2();
-
     typedef db_clock::rep clock_type;
 
     db::all_batches_replayed all_replayed = all_batches_replayed::yes;
@@ -309,26 +172,21 @@ future<db::all_batches_replayed> db::batchlog_manager::replay_all_failed_batches
     auto throttle = _replay_rate / _qp.proxy().get_token_metadata_ptr()->count_normal_token_owners();
     auto limiter = make_lw_shared<utils::rate_limiter>(throttle);
 
-    auto schema = _qp.db().find_schema(system_keyspace::NAME, system_keyspace::BATCHLOG_V2);
-
-    struct replay_stats {
-        std::optional<db_clock::time_point> min_too_fresh;
-        bool need_cleanup = false;
+    auto schema = _qp.db().find_schema(system_keyspace::NAME, system_keyspace::BATCHLOG);
+    auto delete_batch = [this, schema = std::move(schema)] (utils::UUID id) {
+        auto key = partition_key::from_singular(*schema, id);
+        mutation m(schema, key);
+        auto now = service::client_state(service::client_state::internal_tag()).get_timestamp();
+        m.partition().apply_delete(*schema, clustering_key_prefix::make_empty(), tombstone(now, gc_clock::now()));
+        return _qp.proxy().mutate_locally(m, tracing::trace_state_ptr(), db::commitlog::force_sync::no);
     };
 
-    std::unordered_map<int32_t, replay_stats> replay_stats_per_shard;
-
-    // Use a stable `now` accross all batches, so skip/replay decisions are the
-    // same accross a while prefix of written_at (accross all ids).
-    const auto now = db_clock::now();
-
-    auto batch = [this, cleanup, limiter, schema, &all_replayed, &replay_stats_per_shard, now] (const cql3::untyped_result_set::row& row) -> future<stop_iteration> {
-        const auto stage = static_cast<batchlog_stage>(row.get_as<int8_t>("stage"));
-        const auto batch_shard = row.get_as<int32_t>("shard");
+    auto batch = [this, limiter, delete_batch = std::move(delete_batch), &all_replayed](const cql3::untyped_result_set::row& row) -> future<stop_iteration> {
         auto written_at = row.get_as<db_clock::time_point>("written_at");
         auto id = row.get_as<utils::UUID>("id");
         // enough time for the actual write + batchlog entry mutation delivery (two separate requests).
-        auto timeout = _replay_timeout;
+        auto now = db_clock::now();
+        auto timeout = get_batch_log_timeout();
 
         if (utils::get_local_injector().is_enabled("skip_batch_replay")) {
             blogger.debug("Skipping batch replay due to skip_batch_replay injection");
@@ -336,48 +194,52 @@ future<db::all_batches_replayed> db::batchlog_manager::replay_all_failed_batches
             co_return stop_iteration::no;
         }
 
+        // check version of serialization format
+        if (!row.has("version")) {
+            blogger.warn("Skipping logged batch because of unknown version");
+            co_await delete_batch(id);
+            co_return stop_iteration::no;
+        }
+
+        auto version = row.get_as<int32_t>("version");
+        if (version != netw::messaging_service::current_version) {
+            blogger.warn("Skipping logged batch because of incorrect version {}; current version = {}", version, netw::messaging_service::current_version);
+            co_await delete_batch(id);
+            co_return stop_iteration::no;
+        }
+
         auto data = row.get_blob_unfragmented("data");
 
-        blogger.debug("Replaying batch {} from stage {} and batch shard {}", id, int32_t(stage), batch_shard);
-
-        utils::chunked_vector<mutation> mutations;
-        bool send_failed = false;
-
-        auto& shard_written_at = replay_stats_per_shard.try_emplace(batch_shard, replay_stats{}).first->second;
+        blogger.debug("Replaying batch {}", id);
 
         try {
-            utils::chunked_vector<std::pair<canonical_mutation, schema_ptr>> fms;
+            auto fms = make_lw_shared<std::deque<canonical_mutation>>();
             auto in = ser::as_input_stream(data);
             while (in.size()) {
-                auto fm = ser::deserialize(in, std::type_identity<canonical_mutation>());
-                const auto tbl = _qp.db().try_find_table(fm.column_family_id());
-                if (!tbl) {
-                    continue;
-                }
-                if (written_at <= tbl->get_truncation_time()) {
-                    continue;
-                }
-                schema_ptr s = tbl->schema();
-                if (s->tombstone_gc_options().mode() == tombstone_gc_mode::repair) {
-                    timeout = std::min(timeout, std::chrono::duration_cast<db_clock::duration>(s->tombstone_gc_options().propagation_delay_in_seconds()));
-                }
-                fms.emplace_back(std::move(fm), std::move(s));
+                fms->emplace_back(ser::deserialize(in, std::type_identity<canonical_mutation>()));
+                schema_ptr s = _qp.db().find_schema(fms->back().column_family_id());
+                timeout = std::min(timeout, std::chrono::duration_cast<db_clock::duration>(s->tombstone_gc_options().propagation_delay_in_seconds()));
             }
 
             if (now < written_at + timeout) {
                 blogger.debug("Skipping replay of {}, too fresh", id);
-
-                shard_written_at.min_too_fresh = std::min(shard_written_at.min_too_fresh.value_or(written_at), written_at);
-
                 co_return stop_iteration::no;
             }
 
             auto size = data.size();
 
-            for (const auto& [fm, s] : fms) {
-                mutations.emplace_back(fm.to_mutation(s));
-                co_await maybe_yield();
-            }
+            auto mutations = co_await map_reduce(*fms, [this, written_at] (canonical_mutation& fm) {
+                const auto& cf = _qp.proxy().local_db().find_column_family(fm.column_family_id());
+                return make_ready_future<canonical_mutation*>(written_at > cf.get_truncation_time() ? &fm : nullptr);
+            },
+            utils::chunked_vector<mutation>(),
+            [this] (utils::chunked_vector<mutation> mutations, canonical_mutation* fm) {
+                if (fm) {
+                    schema_ptr s = _qp.db().find_schema(fm->column_family_id());
+                    mutations.emplace_back(fm->to_mutation(s));
+                }
+                return mutations;
+            });
 
             if (!mutations.empty()) {
                 const auto ttl = [written_at]() -> clock_type {
@@ -403,11 +265,7 @@ future<db::all_batches_replayed> db::batchlog_manager::replay_all_failed_batches
                     co_await limiter->reserve(size);
                     _stats.write_attempts += mutations.size();
                     auto timeout = db::timeout_clock::now() + write_timeout;
-                    if (cleanup) {
-                        co_await _qp.proxy().send_batchlog_replay_to_all_replicas(mutations, timeout);
-                    } else {
-                        co_await _qp.proxy().send_batchlog_replay_to_all_replicas(std::move(mutations), timeout);
-                    }
+                    co_await _qp.proxy().send_batchlog_replay_to_all_replicas(std::move(mutations), timeout);
                 }
             }
         } catch (data_dictionary::no_such_keyspace& ex) {
@@ -421,80 +279,31 @@ future<db::all_batches_replayed> db::batchlog_manager::replay_all_failed_batches
             // Do _not_ remove the batch, assuning we got a node write error.
             // Since we don't have hints (which origin is satisfied with),
             // we have to resort to keeping this batch to next lap.
-            if (!cleanup || stage == batchlog_stage::failed_replay) {
-                co_return stop_iteration::no;
-            }
-            send_failed = true;
+            co_return stop_iteration::no;
         }
-
-        auto& sp = _qp.proxy();
-
-        if (send_failed) {
-            blogger.debug("Moving batch {} to stage failed_replay", id);
-            auto m = get_batchlog_mutation_for(schema, mutations, netw::messaging_service::current_version, batchlog_stage::failed_replay, written_at, id);
-            co_await sp.mutate_locally(m, tracing::trace_state_ptr(), db::commitlog::force_sync::no);
-        }
-
         // delete batch
-        auto m = get_batchlog_delete_mutation(schema, netw::messaging_service::current_version, stage, written_at, id);
-        co_await _qp.proxy().mutate_locally(m, tracing::trace_state_ptr(), db::commitlog::force_sync::no);
-
-        shard_written_at.need_cleanup = true;
-
+        co_await delete_batch(id);
         co_return stop_iteration::no;
     };
 
-    co_await with_gate(_gate, [this, cleanup, &all_replayed, batch = std::move(batch), now, &replay_stats_per_shard] () mutable -> future<> {
-        blogger.debug("Started replayAllFailedBatches with cleanup: {}", cleanup);
+    co_await with_gate(_gate, [this, cleanup, batch = std::move(batch)] () mutable -> future<> {
+        blogger.debug("Started replayAllFailedBatches (cpu {})", this_shard_id());
         co_await utils::get_local_injector().inject("add_delay_to_batch_replay", std::chrono::milliseconds(1000));
-
-        auto schema = _qp.db().find_schema(system_keyspace::NAME, system_keyspace::BATCHLOG_V2);
-
-        co_await coroutine::parallel_for_each(std::views::iota(0, 16), [&] (int32_t chunk) -> future<> {
-            const int32_t batchlog_chunk_base = chunk * 16;
-            for (int32_t i = 0; i < 16; ++i) {
-                int32_t batchlog_shard = batchlog_chunk_base + i;
-
-                co_await _qp.query_internal(
-                        format("SELECT * FROM {}.{} WHERE version = ? AND stage = ? AND shard = ? BYPASS CACHE", system_keyspace::NAME, system_keyspace::BATCHLOG_V2),
-                        db::consistency_level::ONE,
-                        {data_value(netw::messaging_service::current_version), data_value(int8_t(batchlog_stage::failed_replay)), data_value(batchlog_shard)},
-                        page_size,
-                        batch);
-
-                co_await _qp.query_internal(
-                        format("SELECT * FROM {}.{} WHERE version = ? AND stage = ? AND shard = ? BYPASS CACHE", system_keyspace::NAME, system_keyspace::BATCHLOG_V2),
-                        db::consistency_level::ONE,
-                        {data_value(netw::messaging_service::current_version), data_value(int8_t(batchlog_stage::initial)), data_value(batchlog_shard)},
-                        page_size,
-                        batch);
-
-                if (cleanup != post_replay_cleanup::yes) {
-                    continue;
-                }
-
-                auto it = replay_stats_per_shard.find(batchlog_shard);
-                if (it == replay_stats_per_shard.end() || !it->second.need_cleanup) {
-                    // Nothing was replayed on this batchlog shard, nothing to cleanup.
-                    continue;
-                }
-
-                const auto write_time = it->second.min_too_fresh.value_or(now - _replay_timeout);
-                const auto end_weight  = it->second.min_too_fresh ? bound_weight::before_all_prefixed : bound_weight::after_all_prefixed;
-                auto [key, ckey] = get_batchlog_key(*schema, netw::messaging_service::current_version, batchlog_stage::initial, batchlog_shard, write_time, {});
-                auto end_pos = position_in_partition(partition_region::clustered, end_weight, std::move(ckey));
-
-                range_tombstone rt(position_in_partition::before_all_clustered_rows(), std::move(end_pos), tombstone(api::new_timestamp(), gc_clock::now()));
-
-                blogger.trace("Clean up batchlog shard {} with range tombstone {}", batchlog_shard, rt);
-
-                mutation m(schema, key);
-                m.partition().apply_row_tombstone(*schema, std::move(rt));
-                co_await _qp.proxy().mutate_locally(m, tracing::trace_state_ptr(), db::commitlog::force_sync::no);
+        co_await _qp.query_internal(
+                format("SELECT id, data, written_at, version FROM {}.{} BYPASS CACHE", system_keyspace::NAME, system_keyspace::BATCHLOG),
+                db::consistency_level::ONE,
+                {},
+                page_size,
+                std::move(batch)).then([this, cleanup] {
+            if (cleanup == post_replay_cleanup::no) {
+                return make_ready_future<>();
             }
+            // Replaying batches could have generated tombstones, flush to disk,
+            // where they can be compacted away.
+            return replica::database::flush_table_on_all_shards(_qp.proxy().get_db(), system_keyspace::NAME, system_keyspace::BATCHLOG);
+        }).then([] {
+            blogger.debug("Finished replayAllFailedBatches");
         });
-
-        blogger.debug("Finished replayAllFailedBatches with all_replayed: {}", all_replayed);
     });
 
     co_return all_replayed;

db/batchlog_manager.hh

@@ -34,17 +34,12 @@ class system_keyspace;
 using all_batches_replayed = bool_class<struct all_batches_replayed_tag>;
 
 struct batchlog_manager_config {
-    db_clock::duration replay_timeout;
+    std::chrono::duration<double> write_request_timeout;
     uint64_t replay_rate = std::numeric_limits<uint64_t>::max();
     std::chrono::milliseconds delay = std::chrono::milliseconds(0);
     unsigned replay_cleanup_after_replays;
 };
 
-enum class batchlog_stage : int8_t {
-    initial,
-    failed_replay
-};
-
 class batchlog_manager : public peering_sharded_service<batchlog_manager> {
 public:
     using post_replay_cleanup = bool_class<class post_replay_cleanup_tag>;
@@ -64,7 +59,7 @@ private:
 
     cql3::query_processor& _qp;
     db::system_keyspace& _sys_ks;
-    db_clock::duration _replay_timeout;
+    db_clock::duration _write_request_timeout;
     uint64_t _replay_rate;
     std::chrono::milliseconds _delay;
     unsigned _replay_cleanup_after_replays = 100;
@@ -76,14 +71,6 @@ private:
 
     gc_clock::time_point _last_replay;
 
-    // Was the v1 -> v2 migration already done since last restart?
-    // The migration is attempted once after each restart. This is redundant but
-    // keeps thing simple. Once no upgrade path exists from a ScyllaDB version
-    // which can still produce v1 entries, this migration code can be removed.
-    bool _migration_done = false;
-
-    future<> maybe_migrate_v1_to_v2();
-
     future<all_batches_replayed> replay_all_failed_batches(post_replay_cleanup cleanup);
 public:
     // Takes a QP, not a distributes. Because this object is supposed
@@ -98,13 +85,10 @@ public:
     future<all_batches_replayed> do_batch_log_replay(post_replay_cleanup cleanup);
 
     future<size_t> count_all_batches() const;
+    db_clock::duration get_batch_log_timeout() const;
     gc_clock::time_point get_last_replay() const {
         return _last_replay;
     }
-
-    const stats& stats() const {
-        return _stats;
-    }
 private:
     future<> batchlog_replay_loop();
 };

db/commitlog/commitlog_replayer.cc

@@ -54,14 +54,12 @@ public:
         uint64_t applied_mutations = 0;
         uint64_t corrupt_bytes = 0;
         uint64_t truncated_at = 0;
-        uint64_t broken_files = 0;
 
         stats& operator+=(const stats& s) {
             invalid_mutations += s.invalid_mutations;
             skipped_mutations += s.skipped_mutations;
             applied_mutations += s.applied_mutations;
             corrupt_bytes += s.corrupt_bytes;
-            broken_files += s.broken_files;
             return *this;
         }
         stats operator+(const stats& s) const {
@@ -194,8 +192,6 @@ db::commitlog_replayer::impl::recover(const commitlog::descriptor& d, const comm
             s->corrupt_bytes += e.bytes();
         } catch (commitlog::segment_truncation& e) {
             s->truncated_at = e.position();
-        } catch (commitlog::header_checksum_error&) {
-            ++s->broken_files;
         } catch (...) {
             throw;
         }
@@ -374,9 +370,6 @@ future<> db::commitlog_replayer::recover(std::vector<sstring> files, sstring fna
                     if (stats.truncated_at != 0) {
                         rlogger.warn("Truncated file: {} at position {}.", f, stats.truncated_at);
                     }
-                    if (stats.broken_files != 0) {
-                        rlogger.warn("Corrupted file header: {}. Skipped.", f);
-                    }
                     rlogger.debug("Log replay of {} complete, {} replayed mutations ({} invalid, {} skipped)"
                                     , f
                                     , stats.applied_mutations

db/config.cc

@@ -1152,7 +1152,7 @@ db::config::config(std::shared_ptr<db::extensions> exts)
         "Number of threads with which to deliver hints. In multiple data-center deployments, consider increasing this number because cross data-center handoff is generally slower.")
     , batchlog_replay_throttle_in_kb(this, "batchlog_replay_throttle_in_kb", value_status::Unused, 1024,
         "Total maximum throttle. Throttling is reduced proportionally to the number of nodes in the cluster.")
-    , batchlog_replay_cleanup_after_replays(this, "batchlog_replay_cleanup_after_replays", liveness::LiveUpdate, value_status::Used, 1,
+    , batchlog_replay_cleanup_after_replays(this, "batchlog_replay_cleanup_after_replays", liveness::LiveUpdate, value_status::Used, 60,
         "Clean up batchlog memtable after every N replays. Replays are issued on a timer, every 60 seconds. So if batchlog_replay_cleanup_after_replays is set to 60, the batchlog memtable is flushed every 60 * 60 seconds.")
     /**
     * @Group Request scheduler properties
@@ -1172,17 +1172,6 @@ db::config::config(std::shared_ptr<db::extensions> exts)
         "* default_weight: (Default: 1 **)  How many requests are handled during each turn of the RoundRobin.\n"
         "* weights: (Default: Keyspace: 1)  Takes a list of keyspaces. It sets how many requests are handled during each turn of the RoundRobin, based on the request_scheduler_id.")
     /**
-    * @Group Vector search settings
-    * @GroupDescription Settings for configuring and tuning vector search functionality.
-    */
-    , vector_store_primary_uri(this, "vector_store_primary_uri", liveness::LiveUpdate, value_status::Used, "",
-        "A comma-separated list of primary vector store node URIs. These nodes are preferred for vector search operations.")
-    , vector_store_secondary_uri(this, "vector_store_secondary_uri", liveness::LiveUpdate, value_status::Used, "",
-        "A comma-separated list of secondary vector store node URIs. These nodes are used as a fallback when all primary nodes are unavailable, and are typically located in a different availability zone for high availability.")
-    , vector_store_encryption_options(this, "vector_store_encryption_options", value_status::Used, {},
-        "Options for encrypted connections to the vector store. These options are used for HTTPS URIs in `vector_store_primary_uri` and `vector_store_secondary_uri`. The available options are:\n"
-        "* truststore: (Default: <not set, use system truststore>) Location of the truststore containing the trusted certificate for authenticating remote servers.")
-    /**
     * @Group Security properties
     * @GroupDescription Server and client security settings.
     */
@@ -1470,6 +1459,13 @@ db::config::config(std::shared_ptr<db::extensions> exts)
     , alternator_max_expression_cache_entries_per_shard(this, "alternator_max_expression_cache_entries_per_shard", liveness::LiveUpdate, value_status::Used, 2000, "Maximum number of cached parsed request expressions, per shard.")
     , alternator_max_users_query_size_in_trace_output(this, "alternator_max_users_query_size_in_trace_output", liveness::LiveUpdate, value_status::Used, uint64_t(4096),
             "Maximum size of user's command in trace output (`alternator_op` entry). Larger traces will be truncated and have `<truncated>` message appended - which doesn't count to the maximum limit.")
+    , vector_store_primary_uri(
+              this, "vector_store_primary_uri", liveness::LiveUpdate, value_status::Used, "", "A comma-separated list of primary vector store node URIs. These nodes are preferred for vector search operations.")
+    , vector_store_secondary_uri(this, "vector_store_secondary_uri", liveness::LiveUpdate, value_status::Used, "",
+              "A comma-separated list of secondary vector store node URIs. These nodes are used as a fallback when all primary nodes are unavailable, and are typically located in a different availability zone for high availability.")
+    , vector_store_encryption_options(this, "vector_store_encryption_options", value_status::Used, {},
+        "Options for encrypted connections to the vector store. These options are used for HTTPS URIs in vector_store_primary_uri and vector_store_secondary_uri. The available options are:\n"
+        "* truststore: (Default: <not set. use system truststore>) Location of the truststore containing the trusted certificate for authenticating remote servers.")
     , abort_on_ebadf(this, "abort_on_ebadf", value_status::Used, true, "Abort the server on incorrect file descriptor access. Throws exception when disabled.")
     , sanitizer_report_backtrace(this, "sanitizer_report_backtrace", value_status::Used, false,
             "In debug mode, report log-structured allocator sanitizer violations with a backtrace. Slow.")

db/config.hh

@@ -344,9 +344,6 @@ public:
     named_value<sstring> request_scheduler;
     named_value<sstring> request_scheduler_id;
     named_value<string_map> request_scheduler_options;
-    named_value<sstring> vector_store_primary_uri;
-    named_value<sstring> vector_store_secondary_uri;
-    named_value<string_map> vector_store_encryption_options;
     named_value<sstring> authenticator;
     named_value<sstring> internode_authenticator;
     named_value<sstring> authorizer;
@@ -474,6 +471,10 @@ public:
     named_value<uint32_t> alternator_max_expression_cache_entries_per_shard;
     named_value<uint64_t> alternator_max_users_query_size_in_trace_output;
 
+    named_value<sstring> vector_store_primary_uri;
+    named_value<sstring> vector_store_secondary_uri;
+    named_value<string_map> vector_store_encryption_options;
+
     named_value<bool> abort_on_ebadf;
 
     named_value<bool> sanitizer_report_backtrace;

db/hints/internal/hint_endpoint_manager.cc

@@ -248,7 +248,7 @@ future<db::commitlog> hint_endpoint_manager::add_store() noexcept {
             // which is larger than the segment ID of the RP of the last written hint.
             cfg.base_segment_id = _last_written_rp.base_id();
 
-            return commitlog::create_commitlog(std::move(cfg)).then([this] (this auto, commitlog l) -> future<commitlog> {
+            return commitlog::create_commitlog(std::move(cfg)).then([this] (commitlog l) -> future<commitlog> {
                 // add_store() is triggered every time hint files are forcefully flushed to I/O (every hints_flush_period).
                 // When this happens we want to refill _sender's segments only if it has finished with the segments he had before.
                 if (_sender.have_segments()) {

db/legacy_schema_migrator.cc

@@ -0,0 +1,602 @@
+/*
+ * Modified by ScyllaDB
+ * Copyright (C) 2017-present ScyllaDB
+ */
+
+/*
+ * SPDX-License-Identifier: (LicenseRef-ScyllaDB-Source-Available-1.0 and Apache-2.0)
+ */
+
+// Since Scylla 2.0, we use system tables whose schemas were introduced in
+// Cassandra 3. If Scylla boots to find a data directory with system tables
+// with older schemas - produced by pre-2.0 Scylla or by pre-3.0 Cassandra,
+// we need to migrate these old tables to the new format.
+//
+// We provide here a function, db::legacy_schema_migrator::migrate(),
+// for a one-time migration from old to new system tables. The function
+// reads old system tables, write them back in the new format, and finally
+// delete the old system tables. Scylla's main should call this function and
+// wait for the returned future, before starting to serve the database.
+
+#include <boost/iterator/filter_iterator.hpp>
+#include <seastar/core/future-util.hh>
+#include <seastar/util/log.hh>
+#include <map>
+#include <unordered_set>
+#include <chrono>
+
+#include "replica/database.hh"
+#include "legacy_schema_migrator.hh"
+#include "system_keyspace.hh"
+#include "schema_tables.hh"
+#include "schema/schema_builder.hh"
+#include "service/storage_proxy.hh"
+#include "utils/rjson.hh"
+#include "cql3/query_processor.hh"
+#include "cql3/untyped_result_set.hh"
+#include "cql3/util.hh"
+#include "cql3/statements/property_definitions.hh"
+
+static seastar::logger mlogger("legacy_schema_migrator");
+
+namespace db {
+namespace legacy_schema_migrator {
+
+// local data carriers
+
+class migrator {
+public:
+    static const std::unordered_set<sstring> legacy_schema_tables;
+
+    migrator(sharded<service::storage_proxy>& sp, sharded<replica::database>& db, sharded<db::system_keyspace>& sys_ks, cql3::query_processor& qp)
+                    : _sp(sp), _db(db), _sys_ks(sys_ks), _qp(qp) {
+    }
+    migrator(migrator&&) = default;
+
+    typedef db_clock::time_point time_point;
+
+    // TODO: we don't support triggers.
+    // this is a placeholder.
+    struct trigger {
+        time_point timestamp;
+        sstring name;
+        std::unordered_map<sstring, sstring> options;
+    };
+
+    struct table {
+        time_point timestamp;
+        schema_ptr metadata;
+        std::vector<trigger> triggers;
+    };
+
+    struct type {
+        time_point timestamp;
+        user_type metadata;
+    };
+
+    struct function {
+        time_point timestamp;
+        sstring ks_name;
+        sstring fn_name;
+        std::vector<sstring> arg_names;
+        std::vector<sstring> arg_types;
+        sstring return_type;
+        bool called_on_null_input;
+        sstring language;
+        sstring body;
+    };
+
+    struct aggregate {
+        time_point timestamp;
+        sstring ks_name;
+        sstring fn_name;
+        std::vector<sstring> arg_names;
+        std::vector<sstring> arg_types;
+        sstring return_type;
+        sstring final_func;
+        sstring initcond;
+        sstring state_func;
+        sstring state_type;
+    };
+
+    struct keyspace {
+        time_point timestamp;
+        sstring name;
+        bool durable_writes;
+        std::map<sstring, sstring> replication_params;
+
+        std::vector<table> tables;
+        std::vector<type> types;
+        std::vector<function> functions;
+        std::vector<aggregate> aggregates;
+    };
+
+    class unsupported_feature : public std::runtime_error {
+    public:
+        using runtime_error::runtime_error;
+    };
+
+    static sstring fmt_query(const char* fmt, const char* table) {
+        return fmt::format(fmt::runtime(fmt), db::system_keyspace::NAME, table);
+    }
+
+    typedef ::shared_ptr<cql3::untyped_result_set> result_set_type;
+    typedef const cql3::untyped_result_set::row row_type;
+
+    future<> read_table(keyspace& dst, sstring cf_name, time_point timestamp) {
+        auto fmt = "SELECT * FROM {}.{} WHERE keyspace_name = ? AND columnfamily_name = ?";
+        auto tq = fmt_query(fmt, db::system_keyspace::legacy::COLUMNFAMILIES);
+        auto cq = fmt_query(fmt, db::system_keyspace::legacy::COLUMNS);
+        auto zq = fmt_query(fmt, db::system_keyspace::legacy::TRIGGERS);
+
+        typedef std::tuple<future<result_set_type>, future<result_set_type>, future<result_set_type>, future<db::schema_tables::legacy::schema_mutations>> result_tuple;
+
+        return when_all(_qp.execute_internal(tq, { dst.name, cf_name }, cql3::query_processor::cache_internal::yes),
+                        _qp.execute_internal(cq, { dst.name, cf_name }, cql3::query_processor::cache_internal::yes),
+                        _qp.execute_internal(zq, { dst.name, cf_name }, cql3::query_processor::cache_internal::yes),
+                        db::schema_tables::legacy::read_table_mutations(_sp, dst.name, cf_name, db::system_keyspace::legacy::column_families()))
+                    .then([&dst, cf_name, timestamp](result_tuple&& t) {
+
+            result_set_type tables = std::get<0>(t).get();
+            result_set_type columns = std::get<1>(t).get();
+            result_set_type triggers = std::get<2>(t).get();
+            db::schema_tables::legacy::schema_mutations sm = std::get<3>(t).get();
+
+            row_type& td = tables->one();
+
+            auto ks_name = td.get_as<sstring>("keyspace_name");
+            auto cf_name = td.get_as<sstring>("columnfamily_name");
+            auto id = table_id(td.get_or("cf_id", generate_legacy_id(ks_name, cf_name).uuid()));
+
+            schema_builder builder(dst.name, cf_name, id);
+
+            builder.with_version(sm.digest());
+
+            cf_type cf = sstring_to_cf_type(td.get_or("type", sstring("standard")));
+            if (cf == cf_type::super) {
+                fail(unimplemented::cause::SUPER);
+            }
+
+            auto comparator = td.get_as<sstring>("comparator");
+            bool is_compound = cell_comparator::check_compound(comparator);
+            builder.set_is_compound(is_compound);
+            cell_comparator::read_collections(builder, comparator);
+
+            bool filter_sparse = false;
+
+            data_type default_validator = {};
+            if (td.has("default_validator")) {
+                default_validator = db::schema_tables::parse_type(td.get_as<sstring>("default_validator"));
+                if (default_validator->is_counter()) {
+                    builder.set_is_counter(true);
+                }
+                builder.set_default_validation_class(default_validator);
+            }
+
+            /*
+             * Determine whether or not the table is *really* dense
+             * We cannot trust is_dense value of true (see CASSANDRA-11502, that fixed the issue for 2.2 only, and not retroactively),
+             * but we can trust is_dense value of false.
+             */
+            auto is_dense = td.get_opt<bool>("is_dense");
+            if (!is_dense || *is_dense) {
+                is_dense = [&] {
+                    /*
+                     * As said above, this method is only here because we need to deal with thrift upgrades.
+                     * Once a CF has been "upgraded", i.e. we've rebuilt and save its CQL3 metadata at least once,
+                     * then we'll have saved the "is_dense" value and will be good to go.
+                     *
+                     * But non-upgraded thrift CF (and pre-7744 CF) will have no value for "is_dense", so we need
+                     * to infer that information without relying on it in that case. And for the most part this is
+                     * easy, a CF that has at least one REGULAR definition is not dense. But the subtlety is that not
+                     * having a REGULAR definition may not mean dense because of CQL3 definitions that have only the
+                     * PRIMARY KEY defined.
+                     *
+                     * So we need to recognize those special case CQL3 table with only a primary key. If we have some
+                     * clustering columns, we're fine as said above. So the only problem is that we cannot decide for
+                     * sure if a CF without REGULAR columns nor CLUSTERING_COLUMN definition is meant to be dense, or if it
+                     * has been created in CQL3 by say:
+                     *    CREATE TABLE test (k int PRIMARY KEY)
+                     * in which case it should not be dense. However, we can limit our margin of error by assuming we are
+                     * in the latter case only if the comparator is exactly CompositeType(UTF8Type).
+                     */
+                    std::optional<column_id> max_cl_idx;
+                    const cql3::untyped_result_set::row * regular = nullptr;
+                    for (auto& row : *columns) {
+                        auto kind_str = row.get_as<sstring>("type");
+                        if (kind_str == "compact_value") {
+                            continue;
+                        }
+
+                        auto kind = db::schema_tables::deserialize_kind(kind_str);
+
+                        if (kind == column_kind::regular_column) {
+                            if (regular != nullptr) {
+                                return false;
+                            }
+                            regular = &row;
+                            continue;
+                        }
+                        if (kind == column_kind::clustering_key) {
+                            max_cl_idx = std::max(column_id(row.get_or("component_index", 0)), max_cl_idx.value_or(column_id()));
+                        }
+                    }
+
+                    auto is_cql3_only_pk_comparator = [](const sstring& comparator) {
+                        if (!cell_comparator::check_compound(comparator)) {
+                            return false;
+                        }
+                        // CMH. We don't have composites, nor a parser for it. This is a simple way of c
+                        // checking the same.
+                        auto comma = comparator.find(',');
+                        if (comma != sstring::npos) {
+                            return false;
+                        }
+                        auto off = comparator.find('(');
+                        auto end = comparator.find(')');
+
+                        return comparator.compare(off, end - off, utf8_type->name()) == 0;
+                    };
+
+                    if (max_cl_idx) {
+                        auto n = std::count(comparator.begin(), comparator.end(), ','); // num comp - 1
+                        return *max_cl_idx == n;
+                    }
+
+                    if (regular) {
+                        return false;
+                    }
+
+                    return !is_cql3_only_pk_comparator(comparator);
+
+                }();
+
+                // now, if switched to sparse, remove redundant compact_value column and the last clustering column,
+                // directly copying CASSANDRA-11502 logic. See CASSANDRA-11315.
+
+                filter_sparse = !*is_dense;
+            }
+            builder.set_is_dense(*is_dense);
+
+            auto is_cql = !*is_dense && is_compound;
+            auto is_static_compact = !*is_dense && !is_compound;
+
+            // org.apache.cassandra.schema.LegacySchemaMigrator#isEmptyCompactValueColumn
+            auto is_empty_compact_value = [](const cql3::untyped_result_set::row& column_row) {
+                auto kind_str = column_row.get_as<sstring>("type");
+                // Cassandra only checks for "compact_value", but Scylla generates "regular" instead (#2586)
+                return (kind_str == "compact_value" || kind_str == "regular")
+                       && column_row.get_as<sstring>("column_name").empty();
+            };
+
+            for (auto& row : *columns) {
+                auto kind_str = row.get_as<sstring>("type");
+                auto kind = db::schema_tables::deserialize_kind(kind_str);
+                auto component_index = kind > column_kind::clustering_key ? 0 : column_id(row.get_or("component_index", 0));
+                auto name = row.get_or<sstring>("column_name", sstring());
+                auto validator = db::schema_tables::parse_type(row.get_as<sstring>("validator"));
+
+                if (is_empty_compact_value(row)) {
+                    continue;
+                }
+
+                if (filter_sparse) {
+                    if (kind_str == "compact_value") {
+                        continue;
+                    }
+                    if (kind == column_kind::clustering_key) {
+                        if (cf == cf_type::super && component_index != 0) {
+                            continue;
+                        }
+                        if (cf != cf_type::super && !is_compound) {
+                            continue;
+                        }
+                    }
+                }
+
+                std::optional<index_metadata_kind> index_kind;
+                sstring index_name;
+                index_options_map options;
+                if (row.has("index_type")) {
+                    index_kind = schema_tables::deserialize_index_kind(row.get_as<sstring>("index_type"));
+                }
+                if (row.has("index_name")) {
+                    index_name = row.get_as<sstring>("index_name");
+                }
+                if (row.has("index_options")) {
+                    sstring index_options_str = row.get_as<sstring>("index_options");
+                    options = rjson::parse_to_map<index_options_map>(std::string_view(index_options_str));
+                    sstring type;
+                    auto i = options.find("index_keys");
+                    if (i != options.end()) {
+                        options.erase(i);
+                        type = "KEYS";
+                    }
+                    i = options.find("index_keys_and_values");
+                    if (i != options.end()) {
+                        options.erase(i);
+                        type = "KEYS_AND_VALUES";
+                    }
+                    if (type.empty()) {
+                        if (validator->is_collection() && validator->is_multi_cell()) {
+                            type = "FULL";
+                        } else {
+                            type = "VALUES";
+                        }
+                    }
+                    auto column = cql3::util::maybe_quote(name);
+                    options["target"] = validator->is_collection()
+                                    ? type + "(" + column + ")"
+                                    : column;
+                }
+                if (index_kind) {
+                    // Origin assumes index_name is always set, so let's do the same
+                    builder.with_index(index_metadata(index_name, options, *index_kind, index_metadata::is_local_index::no));
+                }
+
+                data_type column_name_type = [&] {
+                    if (is_static_compact && kind == column_kind::regular_column) {
+                        return db::schema_tables::parse_type(comparator);
+                    }
+                    return utf8_type;
+                }();
+                auto column_name = [&] {
+                    try {
+                        return column_name_type->from_string(name);
+                    } catch (marshal_exception&) {
+                        // #2597: Scylla < 2.0 writes names in serialized form, try to recover
+                        column_name_type->validate(to_bytes_view(name));
+                        return to_bytes(name);
+                    }
+                }();
+                builder.with_column_ordered(column_definition(std::move(column_name), std::move(validator), kind, component_index));
+            }
+
+            if (is_static_compact) {
+                builder.set_regular_column_name_type(db::schema_tables::parse_type(comparator));
+            }
+
+            if (td.has("gc_grace_seconds")) {
+                builder.set_gc_grace_seconds(td.get_as<int32_t>("gc_grace_seconds"));
+            }
+            if (td.has("min_compaction_threshold")) {
+                builder.set_min_compaction_threshold(td.get_as<int32_t>("min_compaction_threshold"));
+            }
+            if (td.has("max_compaction_threshold")) {
+                builder.set_max_compaction_threshold(td.get_as<int32_t>("max_compaction_threshold"));
+            }
+            if (td.has("comment")) {
+                builder.set_comment(td.get_as<sstring>("comment"));
+            }
+            if (td.has("memtable_flush_period_in_ms")) {
+                builder.set_memtable_flush_period(td.get_as<int32_t>("memtable_flush_period_in_ms"));
+            }
+            if (td.has("caching")) {
+                builder.set_caching_options(caching_options::from_sstring(td.get_as<sstring>("caching")));
+            }
+            if (td.has("default_time_to_live")) {
+                builder.set_default_time_to_live(gc_clock::duration(td.get_as<int32_t>("default_time_to_live")));
+            }
+            if (td.has("speculative_retry")) {
+                builder.set_speculative_retry(td.get_as<sstring>("speculative_retry"));
+            }
+            if (td.has("compaction_strategy_class")) {
+                auto strategy = td.get_as<sstring>("compaction_strategy_class");
+                try {
+                    builder.set_compaction_strategy(compaction::compaction_strategy::type(strategy));
+                } catch (const exceptions::configuration_exception& e) {
+                    // If compaction strategy class isn't supported, fallback to incremental.
+                    mlogger.warn("Falling back to incremental compaction strategy after the problem: {}", e.what());
+                    builder.set_compaction_strategy(compaction::compaction_strategy_type::incremental);
+                }
+            }
+            if (td.has("compaction_strategy_options")) {
+                sstring strategy_options_str = td.get_as<sstring>("compaction_strategy_options");
+                builder.set_compaction_strategy_options(rjson::parse_to_map<std::map<sstring, sstring>>(std::string_view(strategy_options_str)));
+            }
+            auto comp_param = td.get_as<sstring>("compression_parameters");
+            compression_parameters cp(rjson::parse_to_map<std::map<sstring, sstring>>(std::string_view(comp_param)));
+            builder.set_compressor_params(cp);
+
+            if (td.has("min_index_interval")) {
+                builder.set_min_index_interval(td.get_as<int32_t>("min_index_interval"));
+            } else if (td.has("index_interval")) { // compatibility
+                builder.set_min_index_interval(td.get_as<int32_t>("index_interval"));
+            }
+            if (td.has("max_index_interval")) {
+                builder.set_max_index_interval(td.get_as<int32_t>("max_index_interval"));
+            }
+            if (td.has("bloom_filter_fp_chance")) {
+                builder.set_bloom_filter_fp_chance(td.get_as<double>("bloom_filter_fp_chance"));
+            } else {
+                builder.set_bloom_filter_fp_chance(builder.get_bloom_filter_fp_chance());
+            }
+            if (td.has("dropped_columns")) {
+                auto map = td.get_map<sstring, int64_t>("dropped_columns");
+                for (auto&& e : map) {
+                    builder.without_column(e.first, api::timestamp_type(e.second));
+                };
+            }
+
+            // ignore version. we're transient
+            if (!triggers->empty()) {
+                throw unsupported_feature("triggers");
+            }
+
+            dst.tables.emplace_back(table{timestamp, builder.build() });
+        });
+    }
+
+    future<> read_tables(keyspace& dst) {
+        auto query = fmt_query("SELECT columnfamily_name, writeTime(type) AS timestamp FROM {}.{} WHERE keyspace_name = ?",
+                        db::system_keyspace::legacy::COLUMNFAMILIES);
+        return _qp.execute_internal(query, {dst.name}, cql3::query_processor::cache_internal::yes).then([this, &dst](result_set_type result) {
+            return parallel_for_each(*result, [this, &dst](row_type& row) {
+                return read_table(dst, row.get_as<sstring>("columnfamily_name"), row.get_as<time_point>("timestamp"));
+            }).finally([result] {});
+        });
+    }
+
+    future<time_point> read_type_timestamp(keyspace& dst, sstring type_name) {
+        // TODO: Unfortunately there is not a single REGULAR column in system.schema_usertypes, so annoyingly we cannot
+        // use the writeTime() CQL function, and must resort to a lower level.
+        // Origin digs up the actual cells of target partition and gets timestamp from there.
+        // We should do the same, but g-dam that's messy. Lets give back dung value for now.
+        return make_ready_future<time_point>(dst.timestamp);
+    }
+
+    future<> read_types(keyspace& dst) {
+        auto query = fmt_query("SELECT * FROM {}.{} WHERE keyspace_name = ?", db::system_keyspace::legacy::USERTYPES);
+        return _qp.execute_internal(query, {dst.name}, cql3::query_processor::cache_internal::yes).then([this, &dst](result_set_type result) {
+            return parallel_for_each(*result, [this, &dst](row_type& row) {
+                auto name = row.get_blob_unfragmented("type_name");
+                auto columns = row.get_list<bytes>("field_names");
+                auto types = row.get_list<sstring>("field_types");
+                std::vector<data_type> field_types;
+                for (auto&& value : types) {
+                    field_types.emplace_back(db::schema_tables::parse_type(value));
+                }
+                auto ut = user_type_impl::get_instance(dst.name, name, columns, field_types, false);
+                return read_type_timestamp(dst, value_cast<sstring>(utf8_type->deserialize(name))).then([ut = std::move(ut), &dst](time_point timestamp) {
+                    dst.types.emplace_back(type{timestamp, ut});
+                });
+            }).finally([result] {});
+        });
+    }
+
+    future<> read_functions(keyspace& dst) {
+        auto query = fmt_query("SELECT * FROM {}.{} WHERE keyspace_name = ?", db::system_keyspace::legacy::FUNCTIONS);
+        return _qp.execute_internal(query, {dst.name}, cql3::query_processor::cache_internal::yes).then([](result_set_type result) {
+            if (!result->empty()) {
+                throw unsupported_feature("functions");
+            }
+        });
+    }
+
+    future<> read_aggregates(keyspace& dst) {
+        auto query = fmt_query("SELECT * FROM {}.{} WHERE keyspace_name = ?", db::system_keyspace::legacy::AGGREGATES);
+        return _qp.execute_internal(query, {dst.name}, cql3::query_processor::cache_internal::yes).then([](result_set_type result) {
+            if (!result->empty()) {
+                throw unsupported_feature("aggregates");
+            }
+        });
+    }
+
+    future<keyspace> read_keyspace(sstring ks_name, bool durable_writes, sstring strategy_class, sstring strategy_options, time_point timestamp) {
+        auto map = rjson::parse_to_map<std::map<sstring, sstring>>(std::string_view(strategy_options));
+        map.emplace("class", std::move(strategy_class));
+        auto ks = ::make_lw_shared<keyspace>(keyspace{timestamp, std::move(ks_name), durable_writes, std::move(map) });
+
+        return read_tables(*ks).then([this, ks] {
+            //Collection<Type> types = readTypes(keyspaceName);
+            return read_types(*ks);
+        }).then([this, ks] {
+            return read_functions(*ks);
+        }).then([this, ks] {
+            return read_aggregates(*ks);
+        }).then([ks] {
+            return make_ready_future<keyspace>(std::move(*ks));
+        });
+    }
+
+    future<> read_all_keyspaces() {
+        static auto ks_filter = [](row_type& row) {
+            auto ks_name = row.get_as<sstring>("keyspace_name");
+            return ks_name != db::system_keyspace::NAME && ks_name != db::schema_tables::v3::NAME;
+        };
+
+        auto query = fmt_query("SELECT keyspace_name, durable_writes, strategy_options, strategy_class, writeTime(durable_writes) AS timestamp FROM {}.{}",
+                        db::system_keyspace::legacy::KEYSPACES);
+
+        return _qp.execute_internal(query, cql3::query_processor::cache_internal::yes).then([this](result_set_type result) {
+            auto i = boost::make_filter_iterator(ks_filter, result->begin(), result->end());
+            auto e = boost::make_filter_iterator(ks_filter, result->end(), result->end());
+            return parallel_for_each(i, e, [this](row_type& row) {
+                return read_keyspace(row.get_as<sstring>("keyspace_name")
+                                , row.get_as<bool>("durable_writes")
+                                , row.get_as<sstring>("strategy_class")
+                                , row.get_as<sstring>("strategy_options")
+                                , row.get_as<db_clock::time_point>("timestamp")
+                                ).then([this](keyspace ks) {
+                    _keyspaces.emplace_back(std::move(ks));
+                   });
+            }).finally([result] {});
+        });
+    }
+
+    future<> drop_legacy_tables() {
+        mlogger.info("Dropping legacy schema tables");
+        auto with_snapshot = !_keyspaces.empty();
+        for (const sstring& cfname : legacy_schema_tables) {
+            co_await replica::database::legacy_drop_table_on_all_shards(_db, _sys_ks, db::system_keyspace::NAME, cfname, with_snapshot);
+        }
+    }
+
+    future<> store_keyspaces_in_new_schema_tables() {
+        mlogger.info("Moving {} keyspaces from legacy schema tables to the new schema keyspace ({})",
+                        _keyspaces.size(), db::schema_tables::v3::NAME);
+
+        utils::chunked_vector<mutation> mutations;
+
+        for (auto& ks : _keyspaces) {
+            auto ksm = ::make_lw_shared<keyspace_metadata>(ks.name
+                            , ks.replication_params["class"] // TODO, make ksm like c3?
+                            , cql3::statements::property_definitions::to_extended_map(ks.replication_params)
+                            , std::nullopt
+                            , std::nullopt
+                            , ks.durable_writes);
+
+            // we want separate time stamps for tables/types, so cannot bulk them into the ksm.
+            for (auto&& m : db::schema_tables::make_create_keyspace_mutations(schema_features::full(), ksm, ks.timestamp.time_since_epoch().count(), false)) {
+                mutations.emplace_back(std::move(m));
+            }
+            for (auto& t : ks.tables) {
+                db::schema_tables::add_table_or_view_to_schema_mutation(t.metadata, t.timestamp.time_since_epoch().count(), true, mutations);
+            }
+            for (auto& t : ks.types) {
+                db::schema_tables::add_type_to_schema_mutation(t.metadata, t.timestamp.time_since_epoch().count(), mutations);
+            }
+        }
+        return _qp.proxy().mutate_locally(std::move(mutations), tracing::trace_state_ptr());
+    }
+
+    future<> flush_schemas() {
+        auto& db = _qp.db().real_database().container();
+        return replica::database::flush_tables_on_all_shards(db, db::schema_tables::all_table_infos(schema_features::full()));
+    }
+
+    future<> migrate() {
+        return read_all_keyspaces().then([this]() {
+            // write metadata to the new schema tables
+            return store_keyspaces_in_new_schema_tables()
+                                                .then(std::bind(&migrator::flush_schemas, this))
+                                                .then(std::bind(&migrator::drop_legacy_tables, this))
+                                                .then([] { mlogger.info("Completed migration of legacy schema tables"); });
+        });
+    }
+
+    sharded<service::storage_proxy>& _sp;
+    sharded<replica::database>& _db;
+    sharded<db::system_keyspace>& _sys_ks;
+    cql3::query_processor& _qp;
+    std::vector<keyspace> _keyspaces;
+};
+
+const std::unordered_set<sstring> migrator::legacy_schema_tables = {
+                db::system_keyspace::legacy::KEYSPACES,
+                db::system_keyspace::legacy::COLUMNFAMILIES,
+                db::system_keyspace::legacy::COLUMNS,
+                db::system_keyspace::legacy::TRIGGERS,
+                db::system_keyspace::legacy::USERTYPES,
+                db::system_keyspace::legacy::FUNCTIONS,
+                db::system_keyspace::legacy::AGGREGATES,
+};
+
+}
+}
+
+future<>
+db::legacy_schema_migrator::migrate(sharded<service::storage_proxy>& sp, sharded<replica::database>& db, sharded<db::system_keyspace>& sys_ks, cql3::query_processor& qp) {
+    return do_with(migrator(sp, db, sys_ks, qp), std::bind(&migrator::migrate, std::placeholders::_1));
+}
+

db/legacy_schema_migrator.hh

@@ -0,0 +1,37 @@
+/*
+ * Modified by ScyllaDB
+ * Copyright (C) 2017-present ScyllaDB
+ */
+
+/*
+ * SPDX-License-Identifier: (LicenseRef-ScyllaDB-Source-Available-1.0 and Apache-2.0)
+ */
+
+#pragma once
+
+#include <seastar/core/future.hh>
+#include <seastar/core/sharded.hh>
+
+#include "seastarx.hh"
+
+namespace replica {
+class database;
+}
+
+namespace cql3 {
+class query_processor;
+}
+
+namespace service {
+class storage_proxy;
+}
+
+namespace db {
+class system_keyspace;
+
+namespace legacy_schema_migrator {
+
+future<> migrate(sharded<service::storage_proxy>&, sharded<replica::database>& db, sharded<db::system_keyspace>& sys_ks, cql3::query_processor&);
+
+}
+}

db/object_storage_endpoint_param.cc

@@ -135,5 +135,5 @@ const std::string db::object_storage_endpoint_param::gs_type = "gs";
 
 auto fmt::formatter<db::object_storage_endpoint_param>::format(const db::object_storage_endpoint_param& e, fmt::format_context& ctx) const
     -> decltype(ctx.out()) {
-    return fmt::format_to(ctx.out(), "object_storage_endpoint_param{}", e.to_json_string());
+    return fmt::format_to(ctx.out(), "object_storage_endpoint_param{{}}", e.to_json_string());
 }

db/partition_snapshot_row_cursor.hh

@@ -542,7 +542,6 @@ public:
     // Returns the range tombstone for the key range adjacent to the cursor's position from the side of smaller keys.
     // Excludes the range for the row itself. That information is returned by range_tombstone_for_row().
     // It's possible that range_tombstone() is empty and range_tombstone_for_row() is not empty.
-    // Note that this is different from the meaning of rows_entry::range_tombstone(), which includes the row itself.
     tombstone range_tombstone() const { return _range_tombstone; }
 
     // Can be called when cursor is pointing at a row.

db/row_cache.cc

@@ -1287,14 +1287,14 @@ row_cache::row_cache(schema_ptr s, snapshot_source src, cache_tracker& tracker,
     , _partitions(dht::raw_token_less_comparator{})
     , _underlying(src())
     , _snapshot_source(std::move(src))
-    , _update_section(abstract_formatter([this] (fmt::context& ctx) {
-        fmt::format_to(ctx.out(), "cache.update {}.{}", _schema->ks_name(), _schema->cf_name());
+    , _update_section(logalloc::allocating_section_namer([this] (fmt::memory_buffer& buf) {
+        fmt::format_to(std::back_inserter(buf), "{}.{}_update", _schema->ks_name(), _schema->cf_name());
     }))
-    , _populate_section(abstract_formatter([this] (fmt::context& ctx) {
-        fmt::format_to(ctx.out(), "cache.populate {}.{}", _schema->ks_name(), _schema->cf_name());
+    , _populate_section(logalloc::allocating_section_namer([this] (fmt::memory_buffer& buf) {
+        fmt::format_to(std::back_inserter(buf), "{}.{}_populate", _schema->ks_name(), _schema->cf_name());
     }))
-    , _read_section(abstract_formatter([this] (fmt::context& ctx) {
-        fmt::format_to(ctx.out(), "cache.read {}.{}", _schema->ks_name(), _schema->cf_name());
+    , _read_section(logalloc::allocating_section_namer([this] (fmt::memory_buffer& buf) {
+        fmt::format_to(std::back_inserter(buf), "{}.{}_read", _schema->ks_name(), _schema->cf_name());
     }))
 {
   try {

db/schema_applier.cc

@@ -1262,9 +1262,16 @@ static future<> do_merge_schema(sharded<service::storage_proxy>& proxy,  sharded
 {
     slogger.trace("do_merge_schema: {}", mutations);
     schema_applier ap(proxy, ss, sys_ks, reload);
-    co_await execute_do_merge_schema(proxy, ap, std::move(mutations)).finally([&ap]() {
-        return ap.destroy();
-    });
+    std::exception_ptr ex;
+    try {
+        co_await execute_do_merge_schema(proxy, ap, std::move(mutations));
+    } catch (...) {
+        ex = std::current_exception();
+    }
+    co_await ap.destroy();
+    if (ex) {
+        throw ex;
+    }
 }
 
 /**

db/schema_tables.cc

@@ -404,7 +404,10 @@ const std::unordered_set<table_id>& schema_tables_holding_schema_mutations() {
                 computed_columns(),
                 dropped_columns(),
                 indexes(),
-                scylla_tables()}) {
+                scylla_tables(),
+                db::system_keyspace::legacy::column_families(),
+                db::system_keyspace::legacy::columns(),
+                db::system_keyspace::legacy::triggers()}) {
             SCYLLA_ASSERT(s->clustering_key_size() > 0);
             auto&& first_column_name = s->clustering_column_at(0).name_as_text();
             SCYLLA_ASSERT(first_column_name == "table_name"
@@ -2837,6 +2840,26 @@ void check_no_legacy_secondary_index_mv_schema(replica::database& db, const view
 }
 
 
+namespace legacy {
+
+table_schema_version schema_mutations::digest() const {
+    md5_hasher h;
+    const db::schema_features no_features;
+    db::schema_tables::feed_hash_for_schema_digest(h, _columnfamilies, no_features);
+    db::schema_tables::feed_hash_for_schema_digest(h, _columns, no_features);
+    return table_schema_version(utils::UUID_gen::get_name_UUID(h.finalize()));
+}
+
+future<schema_mutations> read_table_mutations(sharded<service::storage_proxy>& proxy,
+    sstring keyspace_name, sstring table_name, schema_ptr s)
+{
+    mutation cf_m = co_await read_schema_partition_for_table(proxy, s, keyspace_name, table_name);
+    mutation col_m = co_await read_schema_partition_for_table(proxy, db::system_keyspace::legacy::columns(), keyspace_name, table_name);
+    co_return schema_mutations{std::move(cf_m), std::move(col_m)};
+}
+
+} // namespace legacy
+
 static auto GET_COLUMN_MAPPING_QUERY = format("SELECT column_name, clustering_order, column_name_bytes, kind, position, type FROM system.{} WHERE cf_id = ? AND schema_version = ?",
     db::schema_tables::SCYLLA_TABLE_SCHEMA_HISTORY);
 

db/schema_tables.hh

@@ -155,6 +155,24 @@ schema_ptr scylla_table_schema_history();
 const std::unordered_set<table_id>& schema_tables_holding_schema_mutations();
 }
 
+namespace legacy {
+
+class schema_mutations {
+    mutation _columnfamilies;
+    mutation _columns;
+public:
+    schema_mutations(mutation columnfamilies, mutation columns)
+        : _columnfamilies(std::move(columnfamilies))
+        , _columns(std::move(columns))
+    { }
+    table_schema_version digest() const;
+};
+
+future<schema_mutations> read_table_mutations(sharded<service::storage_proxy>& proxy,
+    sstring keyspace_name, sstring table_name, schema_ptr s);
+
+}
+
 struct qualified_name {
     sstring keyspace_name;
     sstring table_name;

db/system_keyspace.cc

@@ -110,7 +110,6 @@ namespace {
             system_keyspace::v3::CDC_LOCAL,
             system_keyspace::DICTS,
             system_keyspace::VIEW_BUILDING_TASKS,
-            system_keyspace::CLIENT_ROUTES,
         };
         if (ks_name == system_keyspace::NAME && tables.contains(cf_name)) {
             props.enable_schema_commitlog();
@@ -138,7 +137,6 @@ namespace {
                 system_keyspace::ROLE_PERMISSIONS,
                 system_keyspace::DICTS,
                 system_keyspace::VIEW_BUILDING_TASKS,
-                system_keyspace::CLIENT_ROUTES,
             };
             if (ks_name == system_keyspace::NAME && tables.contains(cf_name)) {
                 props.is_group0_table = true;
@@ -215,30 +213,6 @@ schema_ptr system_keyspace::batchlog() {
     return batchlog;
 }
 
-schema_ptr system_keyspace::batchlog_v2() {
-    static thread_local auto batchlog_v2 = [] {
-        schema_builder builder(generate_legacy_id(NAME, BATCHLOG_V2), NAME, BATCHLOG_V2,
-        // partition key
-        {{"version", int32_type}, {"stage", byte_type}, {"shard", int32_type}},
-        // clustering key
-        {{"written_at", timestamp_type}, {"id", uuid_type}},
-        // regular columns
-        {{"data", bytes_type}},
-        // static columns
-        {},
-        // regular column name type
-        utf8_type,
-        // comment
-        "batches awaiting replay"
-       );
-       builder.set_gc_grace_seconds(0);
-       builder.set_caching_options(caching_options::get_disabled_caching_options());
-       builder.with_hash_version();
-       return builder.build(schema_builder::compact_storage::no);
-    }();
-    return batchlog_v2;
-}
-
 /*static*/ schema_ptr system_keyspace::paxos() {
     static thread_local auto paxos = [] {
         // FIXME: switch to the new schema_builder interface (with_column(...), etc)
@@ -311,7 +285,6 @@ schema_ptr system_keyspace::topology() {
             .with_column("tablet_balancing_enabled", boolean_type, column_kind::static_column)
             .with_column("upgrade_state", utf8_type, column_kind::static_column)
             .with_column("global_requests", set_type_impl::get_instance(timeuuid_type, true), column_kind::static_column)
-            .with_column("paused_rf_change_requests", set_type_impl::get_instance(timeuuid_type, true), column_kind::static_column)
             .set_comment("Current state of topology change machine")
             .with_hash_version()
             .build();
@@ -874,6 +847,8 @@ schema_ptr system_keyspace::corrupt_data() {
     return corrupt_data;
 }
 
+static constexpr auto schema_gc_grace = std::chrono::duration_cast<std::chrono::seconds>(days(7)).count();
+
 /*static*/ schema_ptr system_keyspace::scylla_local() {
     static thread_local auto scylla_local = [] {
         schema_builder builder(generate_legacy_id(NAME, SCYLLA_LOCAL), NAME, SCYLLA_LOCAL,
@@ -1385,6 +1360,289 @@ schema_ptr system_keyspace::role_permissions() {
     return schema;
 }
 
+schema_ptr system_keyspace::legacy::hints() {
+    static thread_local auto schema = [] {
+        schema_builder builder(generate_legacy_id(NAME, HINTS), NAME, HINTS,
+        // partition key
+        {{"target_id", uuid_type}},
+        // clustering key
+        {{"hint_id", timeuuid_type}, {"message_version", int32_type}},
+        // regular columns
+        {{"mutation", bytes_type}},
+        // static columns
+        {},
+        // regular column name type
+        utf8_type,
+        // comment
+        "*DEPRECATED* hints awaiting delivery"
+       );
+       builder.set_gc_grace_seconds(0);
+       builder.set_compaction_strategy(compaction::compaction_strategy_type::incremental);
+       builder.set_compaction_strategy_options({{"enabled", "false"}});
+       builder.with(schema_builder::compact_storage::yes);
+       builder.with_hash_version();
+       return builder.build();
+    }();
+    return schema;
+}
+
+schema_ptr system_keyspace::legacy::batchlog() {
+    static thread_local auto schema = [] {
+        schema_builder builder(generate_legacy_id(NAME, BATCHLOG), NAME, BATCHLOG,
+        // partition key
+        {{"id", uuid_type}},
+        // clustering key
+        {},
+        // regular columns
+        {{"data", bytes_type}, {"version", int32_type}, {"written_at", timestamp_type}},
+        // static columns
+        {},
+        // regular column name type
+        utf8_type,
+        // comment
+        "*DEPRECATED* batchlog entries"
+       );
+       builder.set_gc_grace_seconds(0);
+       builder.set_compaction_strategy(compaction::compaction_strategy_type::incremental);
+       builder.set_compaction_strategy_options({{"min_threshold", "2"}});
+       builder.with(schema_builder::compact_storage::no);
+       builder.with_hash_version();
+       return builder.build();
+    }();
+    return schema;
+}
+
+schema_ptr system_keyspace::legacy::keyspaces() {
+    static thread_local auto schema = [] {
+        schema_builder builder(generate_legacy_id(NAME, KEYSPACES), NAME, KEYSPACES,
+        // partition key
+        {{"keyspace_name", utf8_type}},
+        // clustering key
+        {},
+        // regular columns
+        {
+         {"durable_writes", boolean_type},
+         {"strategy_class", utf8_type},
+         {"strategy_options", utf8_type}
+        },
+        // static columns
+        {},
+        // regular column name type
+        utf8_type,
+        // comment
+        "*DEPRECATED* keyspace definitions"
+       );
+       builder.set_gc_grace_seconds(schema_gc_grace);
+       builder.with(schema_builder::compact_storage::yes);
+       builder.with_hash_version();
+       return builder.build();
+    }();
+    return schema;
+}
+
+schema_ptr system_keyspace::legacy::column_families() {
+    static thread_local auto schema = [] {
+        schema_builder builder(generate_legacy_id(NAME, COLUMNFAMILIES), NAME, COLUMNFAMILIES,
+        // partition key
+        {{"keyspace_name", utf8_type}},
+        // clustering key
+        {{"columnfamily_name", utf8_type}},
+        // regular columns
+        {
+         {"bloom_filter_fp_chance", double_type},
+         {"caching", utf8_type},
+         {"cf_id", uuid_type},
+         {"comment", utf8_type},
+         {"compaction_strategy_class", utf8_type},
+         {"compaction_strategy_options", utf8_type},
+         {"comparator", utf8_type},
+         {"compression_parameters", utf8_type},
+         {"default_time_to_live", int32_type},
+         {"default_validator", utf8_type},
+         {"dropped_columns",  map_type_impl::get_instance(utf8_type, long_type, true)},
+         {"gc_grace_seconds", int32_type},
+         {"is_dense", boolean_type},
+         {"key_validator", utf8_type},
+         {"max_compaction_threshold", int32_type},
+         {"max_index_interval", int32_type},
+         {"memtable_flush_period_in_ms", int32_type},
+         {"min_compaction_threshold", int32_type},
+         {"min_index_interval", int32_type},
+         {"speculative_retry", utf8_type},
+         {"subcomparator", utf8_type},
+         {"type", utf8_type},
+         // The following 4 columns are only present up until 2.1.8 tables
+         {"key_aliases", utf8_type},
+         {"value_alias", utf8_type},
+         {"column_aliases", utf8_type},
+         {"index_interval", int32_type},},
+        // static columns
+        {},
+        // regular column name type
+        utf8_type,
+        // comment
+        "*DEPRECATED* table definitions"
+       );
+       builder.set_gc_grace_seconds(schema_gc_grace);
+       builder.with(schema_builder::compact_storage::no);
+       builder.with_hash_version();
+       return builder.build();
+    }();
+    return schema;
+}
+
+schema_ptr system_keyspace::legacy::columns() {
+    static thread_local auto schema = [] {
+        schema_builder builder(generate_legacy_id(NAME, COLUMNS), NAME, COLUMNS,
+        // partition key
+        {{"keyspace_name", utf8_type}},
+        // clustering key
+        {{"columnfamily_name", utf8_type}, {"column_name", utf8_type}},
+        // regular columns
+        {
+            {"component_index", int32_type},
+            {"index_name", utf8_type},
+            {"index_options", utf8_type},
+            {"index_type", utf8_type},
+            {"type", utf8_type},
+            {"validator", utf8_type},
+        },
+        // static columns
+        {},
+        // regular column name type
+        utf8_type,
+        // comment
+        "column definitions"
+        );
+        builder.set_gc_grace_seconds(schema_gc_grace);
+        builder.with(schema_builder::compact_storage::no);
+        builder.with_hash_version();
+        return builder.build();
+    }();
+    return schema;
+}
+
+schema_ptr system_keyspace::legacy::triggers() {
+    static thread_local auto schema = [] {
+        schema_builder builder(generate_legacy_id(NAME, TRIGGERS), NAME, TRIGGERS,
+        // partition key
+        {{"keyspace_name", utf8_type}},
+        // clustering key
+        {{"columnfamily_name", utf8_type}, {"trigger_name", utf8_type}},
+        // regular columns
+        {
+            {"trigger_options",  map_type_impl::get_instance(utf8_type, utf8_type, true)},
+        },
+        // static columns
+        {},
+        // regular column name type
+        utf8_type,
+        // comment
+        "trigger definitions"
+        );
+        builder.set_gc_grace_seconds(schema_gc_grace);
+        builder.with(schema_builder::compact_storage::no);
+        builder.with_hash_version();
+        return builder.build();
+    }();
+    return schema;
+}
+
+schema_ptr system_keyspace::legacy::usertypes() {
+    static thread_local auto schema = [] {
+        schema_builder builder(generate_legacy_id(NAME, USERTYPES), NAME, USERTYPES,
+        // partition key
+        {{"keyspace_name", utf8_type}},
+        // clustering key
+        {{"type_name", utf8_type}},
+        // regular columns
+        {
+            {"field_names", list_type_impl::get_instance(utf8_type, true)},
+            {"field_types", list_type_impl::get_instance(utf8_type, true)},
+        },
+        // static columns
+        {},
+        // regular column name type
+        utf8_type,
+        // comment
+        "user defined type definitions"
+        );
+        builder.set_gc_grace_seconds(schema_gc_grace);
+        builder.with(schema_builder::compact_storage::no);
+        builder.with_hash_version();
+        return builder.build();
+    }();
+    return schema;
+}
+
+schema_ptr system_keyspace::legacy::functions() {
+    /**
+     * Note: we have our own "legacy" version of this table (in schema_tables),
+     * but it is (afaik) not used, and differs slightly from the origin one.
+     * This is based on the origin schema, since we're more likely to encounter
+     * installations of that to migrate, rather than our own (if we dont use the table).
+     */
+    static thread_local auto schema = [] {
+        schema_builder builder(generate_legacy_id(NAME, FUNCTIONS), NAME, FUNCTIONS,
+        // partition key
+        {{"keyspace_name", utf8_type}},
+        // clustering key
+        {{"function_name", utf8_type},{"signature", list_type_impl::get_instance(utf8_type, false)}},
+        // regular columns
+        {
+            {"argument_names", list_type_impl::get_instance(utf8_type, true)},
+            {"argument_types", list_type_impl::get_instance(utf8_type, true)},
+            {"body", utf8_type},
+            {"language", utf8_type},
+            {"return_type", utf8_type},
+            {"called_on_null_input", boolean_type},
+        },
+        // static columns
+        {},
+        // regular column name type
+        utf8_type,
+        // comment
+        "*DEPRECATED* user defined type definitions"
+        );
+        builder.set_gc_grace_seconds(schema_gc_grace);
+        builder.with(schema_builder::compact_storage::no);
+        builder.with_hash_version();
+        return builder.build();
+    }();
+    return schema;
+}
+
+schema_ptr system_keyspace::legacy::aggregates() {
+    static thread_local auto schema = [] {
+        schema_builder builder(generate_legacy_id(NAME, AGGREGATES), NAME, AGGREGATES,
+        // partition key
+        {{"keyspace_name", utf8_type}},
+        // clustering key
+        {{"aggregate_name", utf8_type},{"signature", list_type_impl::get_instance(utf8_type, false)}},
+        // regular columns
+        {
+            {"argument_types", list_type_impl::get_instance(utf8_type, true)},
+            {"final_func", utf8_type},
+            {"initcond", bytes_type},
+            {"return_type", utf8_type},
+            {"state_func", utf8_type},
+            {"state_type", utf8_type},
+        },
+        // static columns
+        {},
+        // regular column name type
+        utf8_type,
+        // comment
+        "*DEPRECATED* user defined aggregate definition"
+        );
+        builder.set_gc_grace_seconds(schema_gc_grace);
+        builder.with(schema_builder::compact_storage::no);
+        builder.with_hash_version();
+        return builder.build();
+    }();
+    return schema;
+}
+
 schema_ptr system_keyspace::dicts() {
     static thread_local auto schema = [] {
         auto id = generate_legacy_id(NAME, DICTS);
@@ -1418,23 +1676,6 @@ schema_ptr system_keyspace::view_building_tasks() {
     return schema;
 }
 
-schema_ptr system_keyspace::client_routes() {
-    static thread_local auto schema = [] {
-        auto id = generate_legacy_id(NAME, CLIENT_ROUTES);
-        return schema_builder(NAME, CLIENT_ROUTES, std::make_optional(id))
-                .with_column("connection_id", utf8_type, column_kind::partition_key)
-                .with_column("host_id", uuid_type, column_kind::clustering_key)
-                .with_column("address", utf8_type)
-                .with_column("port", int32_type)
-                .with_column("tls_port", int32_type)
-                .with_column("alternator_port", int32_type)
-                .with_column("alternator_https_port", int32_type)
-                .with_hash_version()
-                .build();
-    }();
-    return schema;
-}
-
 future<system_keyspace::local_info> system_keyspace::load_local_info() {
     auto msg = co_await execute_cql(format("SELECT host_id, cluster_name, data_center, rack FROM system.{} WHERE key=?", LOCAL), sstring(LOCAL));
 
@@ -2348,7 +2589,7 @@ std::vector<schema_ptr> system_keyspace::all_tables(const db::config& cfg) {
     std::copy(schema_tables.begin(), schema_tables.end(), std::back_inserter(r));
     auto auth_tables = system_keyspace::auth_tables();
     std::copy(auth_tables.begin(), auth_tables.end(), std::back_inserter(r));
-    r.insert(r.end(), { built_indexes(), hints(), batchlog(), batchlog_v2(), paxos(), local(),
+    r.insert(r.end(), { built_indexes(), hints(), batchlog(), paxos(), local(),
                     peers(), peer_events(), range_xfers(),
                     compactions_in_progress(), compaction_history(),
                     sstable_activity(), size_estimates(), large_partitions(), large_rows(), large_cells(),
@@ -2362,7 +2603,7 @@ std::vector<schema_ptr> system_keyspace::all_tables(const db::config& cfg) {
                     v3::cdc_local(),
                     raft(), raft_snapshots(), raft_snapshot_config(), group0_history(), discovery(),
                     topology(), cdc_generations_v3(), topology_requests(), service_levels_v2(), view_build_status_v2(),
-                    dicts(), view_building_tasks(), client_routes(), cdc_streams_state(), cdc_streams_history()
+                    dicts(), view_building_tasks(), cdc_streams_state(), cdc_streams_history()
     });
 
     if (cfg.check_experimental(db::experimental_features_t::feature::BROADCAST_TABLES)) {
@@ -2374,14 +2615,19 @@ std::vector<schema_ptr> system_keyspace::all_tables(const db::config& cfg) {
     if (cfg.check_experimental(db::experimental_features_t::feature::KEYSPACE_STORAGE_OPTIONS)) {
         r.insert(r.end(), {sstables_registry()});
     }
+    // legacy schema
+    r.insert(r.end(), {
+                    // TODO: once we migrate hints/batchlog and add converter
+                    // legacy::hints(), legacy::batchlog(),
+                    legacy::keyspaces(), legacy::column_families(),
+                    legacy::columns(), legacy::triggers(), legacy::usertypes(),
+                    legacy::functions(), legacy::aggregates(), });
 
     return r;
 }
 
 static bool maybe_write_in_user_memory(schema_ptr s) {
-    return (s.get() == system_keyspace::batchlog().get())
-            || (s.get() == system_keyspace::batchlog_v2().get())
-            || (s.get() == system_keyspace::paxos().get())
+    return (s.get() == system_keyspace::batchlog().get()) || (s.get() == system_keyspace::paxos().get())
             || s == system_keyspace::v3::scylla_views_builds_in_progress();
 }
 
@@ -3157,10 +3403,7 @@ static bool must_have_tokens(service::node_state nst) {
     // A decommissioning node doesn't have tokens at the end, they are
     // removed during transition to the left_token_ring state.
     case service::node_state::decommissioning: return false;
-    // A removing node might or might not have tokens depending on whether
-    // REMOVENODE_WITH_LEFT_TOKEN_RING feature is enabled. To support both
-    // cases, we allow removing nodes to not have tokens.
-    case service::node_state::removing: return false;
+    case service::node_state::removing: return true;
     case service::node_state::rebuilding: return true;
     case service::node_state::normal: return true;
     case service::node_state::left: return false;
@@ -3400,12 +3643,6 @@ future<service::topology> system_keyspace::load_topology_state(const std::unorde
             }
         }
 
-        if (some_row.has("paused_rf_change_requests")) {
-            for (auto&& v : deserialize_set_column(*topology(), some_row, "paused_rf_change_requests")) {
-                ret.paused_rf_change_requests.insert(value_cast<utils::UUID>(v));
-            }
-        }
-
         if (some_row.has("enabled_features")) {
             ret.enabled_features = decode_features(deserialize_set_column(*topology(), some_row, "enabled_features"));
         }
@@ -3617,43 +3854,35 @@ system_keyspace::topology_requests_entry system_keyspace::topology_request_row_t
     return entry;
 }
 
-future<system_keyspace::topology_requests_entry> system_keyspace::get_topology_request_entry(utils::UUID id) {
-    auto r = co_await get_topology_request_entry_opt(id);
-    if (!r) {
-        on_internal_error(slogger, format("no entry for request id {}", id));
-    }
-    co_return std::move(*r);
-}
-
-future<std::optional<system_keyspace::topology_requests_entry>> system_keyspace::get_topology_request_entry_opt(utils::UUID id) {
+future<system_keyspace::topology_requests_entry> system_keyspace::get_topology_request_entry(utils::UUID id, bool require_entry) {
     auto rs = co_await execute_cql(
         format("SELECT * FROM system.{} WHERE id = {}", TOPOLOGY_REQUESTS, id));
 
     if (!rs || rs->empty()) {
-        co_return std::nullopt;
+        if (require_entry) {
+            on_internal_error(slogger, format("no entry for request id {}", id));
+        } else {
+            co_return topology_requests_entry{
+                .id = utils::null_uuid()
+            };
+        }
     }
 
     const auto& row = rs->one();
     co_return topology_request_row_to_entry(id, row);
 }
 
-future<system_keyspace::topology_requests_entries> system_keyspace::get_topology_request_entries(std::vector<std::variant<service::topology_request, service::global_topology_request>> request_types, db_clock::time_point end_time_limit) {
-    sstring request_types_str = "";
-    bool first = true;
-    for (const auto& rt : request_types) {
-        if (!std::exchange(first, false)) {
-            request_types_str += ", ";
-        }
-        request_types_str += std::visit([] (auto&& arg) { return fmt::format("'{}'", arg); }, rt);
-    }
-
+future<system_keyspace::topology_requests_entries> system_keyspace::get_node_ops_request_entries(db_clock::time_point end_time_limit) {
     // Running requests.
     auto rs_running = co_await execute_cql(
-        format("SELECT * FROM system.{} WHERE done = false AND request_type IN ({}) ALLOW FILTERING", TOPOLOGY_REQUESTS, request_types_str));
+        format("SELECT * FROM system.{} WHERE done = false AND request_type IN ('{}', '{}', '{}', '{}', '{}') ALLOW FILTERING", TOPOLOGY_REQUESTS,
+            service::topology_request::join, service::topology_request::replace, service::topology_request::rebuild, service::topology_request::leave, service::topology_request::remove));
+
 
     // Requests which finished after end_time_limit.
     auto rs_done = co_await execute_cql(
-        format("SELECT * FROM system.{} WHERE end_time > {} AND request_type IN ({}) ALLOW FILTERING", TOPOLOGY_REQUESTS, end_time_limit.time_since_epoch().count(), request_types_str));
+        format("SELECT * FROM system.{} WHERE end_time > {} AND request_type IN ('{}', '{}', '{}', '{}', '{}') ALLOW FILTERING", TOPOLOGY_REQUESTS, end_time_limit.time_since_epoch().count(),
+            service::topology_request::join, service::topology_request::replace, service::topology_request::rebuild, service::topology_request::leave, service::topology_request::remove));
 
     topology_requests_entries m;
     for (const auto& row: *rs_done) {
@@ -3671,16 +3900,6 @@ future<system_keyspace::topology_requests_entries> system_keyspace::get_topology
     co_return m;
 }
 
-future<system_keyspace::topology_requests_entries> system_keyspace::get_node_ops_request_entries(db_clock::time_point end_time_limit) {
-    return get_topology_request_entries({
-        service::topology_request::join,
-        service::topology_request::replace,
-        service::topology_request::rebuild,
-        service::topology_request::leave,
-        service::topology_request::remove
-    }, end_time_limit);
-}
-
 future<mutation> system_keyspace::get_insert_dict_mutation(
     std::string_view name,
     bytes data,

db/system_keyspace.hh

@@ -163,7 +163,6 @@ public:
     static constexpr auto NAME = "system";
     static constexpr auto HINTS = "hints";
     static constexpr auto BATCHLOG = "batchlog";
-    static constexpr auto BATCHLOG_V2 = "batchlog_v2";
     static constexpr auto PAXOS = "paxos";
     static constexpr auto BUILT_INDEXES = "IndexInfo";
     static constexpr auto LOCAL = "local";
@@ -199,7 +198,6 @@ public:
     static constexpr auto VIEW_BUILD_STATUS_V2 = "view_build_status_v2";
     static constexpr auto DICTS = "dicts";
     static constexpr auto VIEW_BUILDING_TASKS = "view_building_tasks";
-    static constexpr auto CLIENT_ROUTES = "client_routes";
 
     // auth
     static constexpr auto ROLES = "roles";
@@ -243,6 +241,28 @@ public:
         static schema_ptr cdc_local();
     };
 
+    struct legacy {
+        static constexpr auto HINTS = "hints";
+        static constexpr auto BATCHLOG = "batchlog";
+        static constexpr auto KEYSPACES = "schema_keyspaces";
+        static constexpr auto COLUMNFAMILIES = "schema_columnfamilies";
+        static constexpr auto COLUMNS = "schema_columns";
+        static constexpr auto TRIGGERS = "schema_triggers";
+        static constexpr auto USERTYPES = "schema_usertypes";
+        static constexpr auto FUNCTIONS = "schema_functions";
+        static constexpr auto AGGREGATES = "schema_aggregates";
+
+        static schema_ptr keyspaces();
+        static schema_ptr column_families();
+        static schema_ptr columns();
+        static schema_ptr triggers();
+        static schema_ptr usertypes();
+        static schema_ptr functions();
+        static schema_ptr aggregates();
+        static schema_ptr hints();
+        static schema_ptr batchlog();
+    };
+
     // Partition estimates for a given range of tokens.
     struct range_estimates {
         schema_ptr schema;
@@ -257,7 +277,6 @@ public:
 
     static schema_ptr hints();
     static schema_ptr batchlog();
-    static schema_ptr batchlog_v2();
     static schema_ptr paxos();
     static schema_ptr built_indexes(); // TODO (from Cassandra): make private
     static schema_ptr raft();
@@ -277,7 +296,6 @@ public:
     static schema_ptr view_build_status_v2();
     static schema_ptr dicts();
     static schema_ptr view_building_tasks();
-    static schema_ptr client_routes();
 
     // auth
     static schema_ptr roles();
@@ -669,9 +687,7 @@ public:
 
     future<service::topology_request_state> get_topology_request_state(utils::UUID id, bool require_entry);
     topology_requests_entry topology_request_row_to_entry(utils::UUID id, const cql3::untyped_result_set_row& row);
-    future<topology_requests_entry> get_topology_request_entry(utils::UUID id);
-    future<std::optional<topology_requests_entry>> get_topology_request_entry_opt(utils::UUID id);
-    future<system_keyspace::topology_requests_entries> get_topology_request_entries(std::vector<std::variant<service::topology_request, service::global_topology_request>> request_types, db_clock::time_point end_time_limit);
+    future<topology_requests_entry> get_topology_request_entry(utils::UUID id, bool require_entry);
     future<topology_requests_entries> get_node_ops_request_entries(db_clock::time_point end_time_limit);
 
 public:

db/view/delete_ghost_rows_visitor.hh

@@ -9,8 +9,6 @@
 #include "query/query-result-reader.hh"
 #include "replica/database_fwd.hh"
 #include "db/timeout_clock.hh"
-#include <seastar/core/future.hh>
-#include <seastar/core/gate.hh>
 
 namespace service {
 class storage_proxy;
@@ -27,14 +25,8 @@ class delete_ghost_rows_visitor {
     replica::table& _view_table;
     schema_ptr _base_schema;
     std::optional<partition_key> _view_pk;
-    db::timeout_semaphore _concurrency_semaphore;
-    seastar::gate _gate;
-    std::exception_ptr& _ex;
-
 public:
-    delete_ghost_rows_visitor(service::storage_proxy& proxy, service::query_state& state, view_ptr view, db::timeout_clock::duration timeout_duration, size_t concurrency, std::exception_ptr& ex);
-    delete_ghost_rows_visitor(delete_ghost_rows_visitor&&) = default;
-    ~delete_ghost_rows_visitor() noexcept;
+    delete_ghost_rows_visitor(service::storage_proxy& proxy, service::query_state& state, view_ptr view, db::timeout_clock::duration timeout_duration);
 
     void add_value(const column_definition& def, query::result_row_view::iterator_type& i) {
     }
@@ -53,9 +45,6 @@ public:
     uint32_t accept_partition_end(const query::result_row_view& static_row) {
         return 0;
     }
-
-private:
-    future<> do_accept_new_row(partition_key pk,  clustering_key ck);
 };
 
 } //namespace db::view

db/view/view.cc

@@ -1744,115 +1744,6 @@ bool should_generate_view_updates_on_this_shard(const schema_ptr& base, const lo
         && std::ranges::contains(shards, this_shard_id());
 }
 
-static endpoints_to_update get_view_natural_endpoint_vnodes(
-        locator::host_id me,
-        std::vector<std::reference_wrapper<const locator::node>> base_nodes,
-        std::vector<std::reference_wrapper<const locator::node>> view_nodes,
-        locator::endpoint_dc_rack my_location,
-        const locator::network_topology_strategy* network_topology,
-        replica::cf_stats& cf_stats) {
-    using node_vector = std::vector<std::reference_wrapper<const locator::node>>;
-    node_vector base_endpoints, view_endpoints;
-    auto& my_datacenter = my_location.dc;
-
-    auto process_candidate = [&] (node_vector& nodes, std::reference_wrapper<const locator::node> node) {
-        if (!network_topology || node.get().dc() == my_datacenter) {
-            nodes.emplace_back(node);
-        }
-    };
-
-    for (auto&& base_node : base_nodes) {
-        process_candidate(base_endpoints, base_node);
-    }
-
-    for (auto&& view_node : view_nodes) {
-        auto it = std::ranges::find(base_endpoints, view_node.get().host_id(), std::mem_fn(&locator::node::host_id));
-        // If this base replica is also one of the view replicas, we use
-        // ourselves as the view replica.
-        // We don't return an extra endpoint, as it's only needed when
-        // using tablets (so !use_legacy_self_pairing)
-        if (view_node.get().host_id() == me && it != base_endpoints.end()) {
-            return {.natural_endpoint = me};
-        }
-
-        // We have to remove any endpoint which is shared between the base
-        // and the view, as it will select itself and throw off the counts
-        // otherwise.
-        if (it != base_endpoints.end()) {
-            base_endpoints.erase(it);
-        } else if (!network_topology || view_node.get().dc() == my_datacenter) {
-            view_endpoints.push_back(view_node);
-        }
-    }
-
-    auto base_it = std::ranges::find(base_endpoints, me, std::mem_fn(&locator::node::host_id));
-    if (base_it == base_endpoints.end()) {
-        // This node is not a base replica of this key, so we return empty
-        // FIXME: This case shouldn't happen, and if it happens, a view update
-        // would be lost.
-        ++cf_stats.total_view_updates_on_wrong_node;
-        vlogger.warn("Could not find {} in base_endpoints={}", me,
-                base_endpoints | std::views::transform(std::mem_fn(&locator::node::host_id)));
-        return {};
-    }
-    size_t idx = base_it - base_endpoints.begin();
-    return {.natural_endpoint = view_endpoints[idx].get().host_id()};
-}
-
-static std::optional<locator::host_id> get_unpaired_view_endpoint(
-        std::vector<std::reference_wrapper<const locator::node>> base_nodes,
-        std::vector<std::reference_wrapper<const locator::node>> view_nodes,
-        replica::cf_stats& cf_stats) {
-    std::unordered_set<locator::endpoint_dc_rack> base_dc_racks;
-    for (auto&& base_node : base_nodes) {
-        if (base_dc_racks.contains(base_node.get().dc_rack())) {
-            // We can't do rack-aware pairing if there are multiple replicas in the same rack.
-            ++cf_stats.total_view_updates_failed_pairing;
-            vlogger.warn("Can't perform base-view pairing in this topology. There are multiple base table replicas in the same dc/rack({}/{}):",
-                    base_node.get().dc(), base_node.get().rack());
-            return std::nullopt;
-        }
-        base_dc_racks.insert(base_node.get().dc_rack());
-    }
-
-    std::unordered_set<locator::endpoint_dc_rack> paired_view_dc_racks;
-    std::unordered_map<locator::endpoint_dc_rack, locator::host_id> unpaired_view_dc_rack_replicas;
-    for (auto&& view_node : view_nodes) {
-        if (paired_view_dc_racks.contains(view_node.get().dc_rack()) || unpaired_view_dc_rack_replicas.contains(view_node.get().dc_rack())) {
-            // We can't do rack-aware pairing if there are multiple replicas in the same rack.
-            ++cf_stats.total_view_updates_failed_pairing;
-            vlogger.warn("Can't perform base-view pairing in this topology. There are multiple view table replicas in the same dc/rack({}/{}):",
-                    view_node.get().dc(), view_node.get().rack());
-            return std::nullopt;
-        }
-        // Track unpaired replicas in both sets
-        if (base_dc_racks.contains(view_node.get().dc_rack())) {
-            paired_view_dc_racks.insert(view_node.get().dc_rack());
-        } else {
-            unpaired_view_dc_rack_replicas.insert({view_node.get().dc_rack(), view_node.get().host_id()});
-        }
-    }
-
-    if (unpaired_view_dc_rack_replicas.size() > 0) {
-        // There are view replicas that can't be paired with any base replica
-        // This can happen as a result of an RF change when the view replica finishes streaming
-        // before the base replica.
-        // Because of this, a view replica might not get paired with any base replica, so we need
-        // to send an additional update to it.
-        ++cf_stats.total_view_updates_due_to_replica_count_mismatch;
-        auto extra_replica = unpaired_view_dc_rack_replicas.begin()->second;
-        unpaired_view_dc_rack_replicas.erase(unpaired_view_dc_rack_replicas.begin());
-        if (unpaired_view_dc_rack_replicas.size() > 0) {
-            // We only expect one extra replica to appear due to an RF change. If there's more, that's an error,
-            // but we'll still perform updates to the paired and last replicas to minimize degradation.
-            vlogger.warn("There are too many view endpoints for base-view pairing. View updates may get lost on view_endpoints={}",
-                    unpaired_view_dc_rack_replicas | std::views::values);
-        }
-        return extra_replica;
-    }
-    return std::nullopt;
-}
-
 // Calculate the node ("natural endpoint") to which this node should send
 // a view update.
 //
@@ -1865,19 +1756,29 @@ static std::optional<locator::host_id> get_unpaired_view_endpoint(
 // of this function is to find, assuming that this node is one of the base
 // replicas for a given partition, the paired view replica.
 //
-// When using vnodes, we have an optimization called "self-pairing" - if a single
-// node is both a base replica and a view replica for a write, the pairing is
-// modified so that this node sends the update to itself and this node is removed
-// from the lists of nodes paired by index. This self-pairing optimization can
-// cause the pairing to change after view ranges are moved between nodes.
+// In the past, we used an optimization called "self-pairing" that if a single
+// node was both a base replica and a view replica for a write, the pairing is
+// modified so that this node would send the update to itself. This self-
+// pairing optimization could cause the pairing to change after view ranges
+// are moved between nodes, so currently we only use it if
+// use_legacy_self_pairing is set to true. When using tablets - where range
+// movements are common - it is strongly recommended to set it to false.
 //
 // If the keyspace's replication strategy is a NetworkTopologyStrategy,
 // we pair only nodes in the same datacenter.
 //
-// If the table uses tablets, then pairing is rack-aware. In this case, in each
-// rack where we have a base replica there is also one replica of each view tablet.
-// Therefore, the base replicas are naturally paired with the view replicas that
-// are in the same rack.
+// When use_legacy_self_pairing is enabled, if one of the base replicas
+// also happens to be a view replica, it is paired with itself
+// (with the other nodes paired by order in the list
+// after taking this node out).
+//
+// If the table uses tablets and the replication strategy is NetworkTopologyStrategy
+// and the replication factor in the node's datacenter is a multiple of the number
+// of racks in the datacenter, then pairing is rack-aware.  In this case,
+// all racks have the same number of replicas, and those are never migrated
+// outside their racks. Therefore, the base replicas are naturally paired with the
+// view replicas that are in the same rack, based on the ordinal position.
+// Note that typically, there is a single replica per rack and pairing is trivial.
 //
 // If the assumption that the given base token belongs to this replica
 // does not hold, we return an empty optional.
@@ -1905,12 +1806,19 @@ endpoints_to_update get_view_natural_endpoint(
         const locator::abstract_replication_strategy& replication_strategy,
         const dht::token& base_token,
         const dht::token& view_token,
-        bool use_tablets,
+        bool use_legacy_self_pairing,
+        bool use_tablets_rack_aware_view_pairing,
         replica::cf_stats& cf_stats) {
     auto& topology = base_erm->get_token_metadata_ptr()->get_topology();
     auto& view_topology = view_erm->get_token_metadata_ptr()->get_topology();
     auto& my_location = topology.get_location(me);
+    auto& my_datacenter = my_location.dc;
     auto* network_topology = dynamic_cast<const locator::network_topology_strategy*>(&replication_strategy);
+    auto rack_aware_pairing = use_tablets_rack_aware_view_pairing && network_topology;
+    bool simple_rack_aware_pairing = false;
+    using node_vector = std::vector<std::reference_wrapper<const locator::node>>;
+    node_vector orig_base_endpoints, orig_view_endpoints;
+    node_vector base_endpoints, view_endpoints;
 
     auto resolve = [&] (const locator::topology& topology, const locator::host_id& ep, bool is_view) -> const locator::node& {
         if (auto* np = topology.find_node(ep)) {
@@ -1921,7 +1829,6 @@ endpoints_to_update get_view_natural_endpoint(
 
     // We need to use get_replicas() for pairing to be stable in case base or view tablet
     // is rebuilding a replica which has left the ring. get_natural_endpoints() filters such replicas.
-    using node_vector = std::vector<std::reference_wrapper<const locator::node>>;
     auto base_nodes = base_erm->get_replicas(base_token) | std::views::transform([&] (const locator::host_id& ep) -> const locator::node& {
         return resolve(topology, ep, false);
     }) | std::ranges::to<node_vector>();
@@ -1945,43 +1852,231 @@ endpoints_to_update get_view_natural_endpoint(
                 // note that the recursive call will not recurse again because leaving_base is in base_nodes.
                 auto leaving_base = it->get().host_id();
                 return get_view_natural_endpoint(leaving_base, base_erm, view_erm, replication_strategy, base_token,
-                        view_token, use_tablets, cf_stats);
+                        view_token, use_legacy_self_pairing, use_tablets_rack_aware_view_pairing, cf_stats);
             }
         }
     }
 
-    if (!use_tablets) {
-        return get_view_natural_endpoint_vnodes(
-                me,
-                base_nodes,
-                view_nodes,
-                my_location,
-                network_topology,
-                cf_stats);
+    std::function<bool(const locator::node&)> is_candidate;
+    if (network_topology) {
+        is_candidate = [&] (const locator::node& node) { return node.dc() == my_datacenter; };
+    } else {
+        is_candidate = [&] (const locator::node&) { return true; };
+    }
+    auto process_candidate = [&] (node_vector& nodes, std::reference_wrapper<const locator::node> node) {
+        if (is_candidate(node)) {
+            nodes.emplace_back(node);
+        }
+    };
+
+    for (auto&& base_node : base_nodes) {
+        process_candidate(base_endpoints, base_node);
     }
 
-    std::optional<locator::host_id> paired_replica;
-    for (auto&& view_node : view_nodes) {
-        if (view_node.get().dc_rack() == my_location) {
-            paired_replica = view_node.get().host_id();
-            break;
+    if (use_legacy_self_pairing) {
+        for (auto&& view_node : view_nodes) {
+            auto it = std::ranges::find(base_endpoints, view_node.get().host_id(), std::mem_fn(&locator::node::host_id));
+            // If this base replica is also one of the view replicas, we use
+            // ourselves as the view replica.
+            // We don't return an extra endpoint, as it's only needed when
+            // using tablets (so !use_legacy_self_pairing)
+            if (view_node.get().host_id() == me && it != base_endpoints.end()) {
+                return {.natural_endpoint = me};
+            }
+
+            // We have to remove any endpoint which is shared between the base
+            // and the view, as it will select itself and throw off the counts
+            // otherwise.
+            if (it != base_endpoints.end()) {
+                base_endpoints.erase(it);
+            } else if (is_candidate(view_node)) {
+                view_endpoints.push_back(view_node);
+            }
+        }
+    } else {
+        for (auto&& view_node : view_nodes) {
+            process_candidate(view_endpoints, view_node);
         }
     }
-    if (paired_replica && base_nodes.size() == view_nodes.size()) {
-        // We don't need to find any extra replicas, so we can return early
-        return {.natural_endpoint = paired_replica};
+
+    // Try optimizing for simple rack-aware pairing
+    // If the numbers of base and view replica differ, that means an RF change is taking place
+    // and we can't use simple rack-aware pairing.
+    if (rack_aware_pairing && base_endpoints.size() == view_endpoints.size()) {
+        auto dc_rf = network_topology->get_replication_factor(my_datacenter);
+        const auto& racks = topology.get_datacenter_rack_nodes().at(my_datacenter);
+        // Simple rack-aware pairing is possible when the datacenter replication factor
+        // is a multiple of the number of racks in the datacenter.
+        if (dc_rf % racks.size() == 0) {
+            simple_rack_aware_pairing = true;
+            size_t rack_rf = dc_rf / racks.size();
+            // If any rack doesn't have enough nodes to satisfy the per-rack rf
+            // simple rack-aware pairing is disabled.
+            for (const auto& [rack, nodes] : racks) {
+                if (nodes.size() < rack_rf) {
+                    simple_rack_aware_pairing = false;
+                    break;
+                }
+            }
+        }
+        if (dc_rf != base_endpoints.size()) {
+            // If the datacenter replication factor is not equal to the number of base replicas,
+            // we're in progress of a RF change and we can't use simple rack-aware pairing.
+            simple_rack_aware_pairing = false;
+        }
+        if (simple_rack_aware_pairing) {
+            std::erase_if(base_endpoints, [&] (const locator::node& node) { return node.dc_rack() != my_location; });
+            std::erase_if(view_endpoints, [&] (const locator::node& node) { return node.dc_rack() != my_location; });
+        }
     }
-    if (!paired_replica) {
-        // We couldn't find any view replica in our rack
+
+    orig_base_endpoints = base_endpoints;
+    orig_view_endpoints = view_endpoints;
+
+    // For the complex rack_aware_pairing case, nodes are already filtered by datacenter
+    // Use best-match, for the minimum number of base and view replicas in each rack,
+    // and ordinal match for the rest.
+    std::optional<std::reference_wrapper<const locator::node>> paired_replica;
+    if (rack_aware_pairing && !simple_rack_aware_pairing) {
+        struct indexed_replica {
+            size_t idx;
+            std::reference_wrapper<const locator::node> node;
+        };
+        std::unordered_map<sstring, std::vector<indexed_replica>> base_racks, view_racks;
+
+        // First, index all replicas by rack
+        auto index_replica_set = [] (std::unordered_map<sstring, std::vector<indexed_replica>>& racks, const node_vector& replicas) {
+            size_t idx = 0;
+            for (const auto& r: replicas) {
+                racks[r.get().rack()].emplace_back(idx++, r);
+            }
+        };
+        index_replica_set(base_racks, base_endpoints);
+        index_replica_set(view_racks, view_endpoints);
+
+        // Try optimistically pairing `me` first
+        const auto& my_base_replicas = base_racks[my_location.rack];
+        auto base_it = std::ranges::find(my_base_replicas, me, [] (const indexed_replica& ir) { return ir.node.get().host_id(); });
+        if (base_it == my_base_replicas.end()) {
+            return {};
+        }
+        const auto& my_view_replicas = view_racks[my_location.rack];
+        size_t idx = base_it - my_base_replicas.begin();
+        if (idx < my_view_replicas.size()) {
+            if (orig_view_endpoints.size() <= orig_base_endpoints.size()) {
+                return {.natural_endpoint = my_view_replicas[idx].node.get().host_id()};
+            } else {
+                // If the number of view replicas is larger than the number of base replicas,
+                // we need to find the unpaired view replica, so we can't return yet.
+                paired_replica = my_view_replicas[idx].node;
+            }
+        }
+
+        // Collect all unpaired base and view replicas,
+        // where the number of replicas in the base rack is different than the respective view rack
+        std::vector<indexed_replica> unpaired_base_replicas, unpaired_view_replicas;
+        for (const auto& [rack, base_replicas] : base_racks) {
+            const auto& view_replicas = view_racks[rack];
+            for (auto i = view_replicas.size(); i < base_replicas.size(); ++i) {
+                unpaired_base_replicas.emplace_back(base_replicas[i]);
+            }
+        }
+        for (const auto& [rack, view_replicas] : view_racks) {
+            const auto& base_replicas = base_racks[rack];
+            for (auto i = base_replicas.size(); i < view_replicas.size(); ++i) {
+                unpaired_view_replicas.emplace_back(view_replicas[i]);
+            }
+        }
+
+        // Sort by the original ordinality, and copy the sorted results
+        // back into {base,view}_endpoints, for backward compatible processing below.
+        std::ranges::sort(unpaired_base_replicas, std::less(), std::mem_fn(&indexed_replica::idx));
+        base_endpoints.clear();
+        std::ranges::transform(unpaired_base_replicas, std::back_inserter(base_endpoints), std::mem_fn(&indexed_replica::node));
+
+        std::ranges::sort(unpaired_view_replicas, std::less(), std::mem_fn(&indexed_replica::idx));
+        view_endpoints.clear();
+        std::ranges::transform(unpaired_view_replicas, std::back_inserter(view_endpoints), std::mem_fn(&indexed_replica::node));
+    }
+
+    auto base_it = std::ranges::find(base_endpoints, me, std::mem_fn(&locator::node::host_id));
+    if (!paired_replica && base_it == base_endpoints.end()) {
+        // This node is not a base replica of this key, so we return empty
+        // FIXME: This case shouldn't happen, and if it happens, a view update
+        // would be lost.
+        ++cf_stats.total_view_updates_on_wrong_node;
+        vlogger.warn("Could not find {} in base_endpoints={}", me,
+                orig_base_endpoints | std::views::transform(std::mem_fn(&locator::node::host_id)));
+        return {};
+    }
+    size_t idx = base_it - base_endpoints.begin();
+    std::optional<std::reference_wrapper<const locator::node>> no_pairing_replica;
+    if (!paired_replica && idx >= view_endpoints.size()) {
+        // There are fewer view replicas than base replicas
+        // FIXME: This might still happen when reducing replication factor with tablets,
+        // see https://github.com/scylladb/scylladb/issues/21492
         ++cf_stats.total_view_updates_failed_pairing;
-        vlogger.warn("Could not find a view replica in the same rack as base replica {} for base_endpoints={} view_endpoints={}", 
-                me,
-                base_nodes | std::views::transform(std::mem_fn(&locator::node::host_id)),
-                view_nodes | std::views::transform(std::mem_fn(&locator::node::host_id)));
+        vlogger.warn("Could not pair {}: rack_aware={} base_endpoints={} view_endpoints={}", me,
+                rack_aware_pairing ? (simple_rack_aware_pairing ? "simple" : "complex") : "none",
+                orig_base_endpoints | std::views::transform(std::mem_fn(&locator::node::host_id)),
+                orig_view_endpoints | std::views::transform(std::mem_fn(&locator::node::host_id)));
+        return {};
+    } else if (base_endpoints.size() < view_endpoints.size()) {
+        // There are fewer base replicas than view replicas.
+        // This can happen as a result of an RF change when the view replica finishes streaming
+        // before the base replica.
+        // Because of this, a view replica might not get paired with any base replica, so we need
+        // to send an additional update to it.
+        ++cf_stats.total_view_updates_due_to_replica_count_mismatch;
+        no_pairing_replica = view_endpoints.back();
+        if (base_endpoints.size() < view_endpoints.size() - 1) {
+            // We only expect one extra replica to appear due to an RF change. If there's more, that's an error,
+            // but we'll still perform updates to the paired and last replicas to minimize degradation.
+            vlogger.warn("There are too many view endpoints for base-view pairing. View updates may get lost on view_endpoints={}",
+                    std::span(view_endpoints.begin() + base_endpoints.size(), view_endpoints.end() - 1) | std::views::transform(std::mem_fn(&locator::node::host_id)));
+        }
     }
-    std::optional<locator::host_id> no_pairing_replica = get_unpaired_view_endpoint(base_nodes, view_nodes, cf_stats);
-    return {.natural_endpoint = paired_replica,
-            .endpoint_with_no_pairing = no_pairing_replica};
+
+    if (!paired_replica) {
+        paired_replica = view_endpoints[idx];
+    }
+    if (!no_pairing_replica && base_nodes.size() < view_nodes.size()) {
+        // This can happen when the view replica with no pairing is in another DC.
+        // We need to send an update to it if there are no base replicas in that DC yet,
+        // as it won't receive updates otherwise.
+        std::unordered_set<sstring> dcs_with_base_replicas;
+        for (const auto& base_node : base_nodes) {
+            dcs_with_base_replicas.insert(base_node.get().dc());
+        }
+        for (const auto& view_node : view_nodes) {
+            if (!dcs_with_base_replicas.contains(view_node.get().dc())) {
+                ++cf_stats.total_view_updates_due_to_replica_count_mismatch;
+                no_pairing_replica = view_node;
+                break;
+            }
+        }
+    }
+    // https://github.com/scylladb/scylladb/issues/19439
+    // With tablets, a node being replaced might transition to "left" state
+    // but still be kept as a replica.
+    // As of writing this hints are not prepared to handle nodes that are left
+    // but are still replicas. Therefore, there is no other sensible option
+    // right now but to give up attempt to send the update or write a hint
+    // to the paired, permanently down replica.
+    // We use the same workaround for the extra replica.
+    auto return_host_id_if_not_left = [] (const auto& replica) -> std::optional<locator::host_id> {
+        if (!replica) {
+            return std::nullopt;
+        }
+        const auto& node = replica->get();
+        if (!node.left()) {
+            return node.host_id();
+        } else {
+            return std::nullopt;
+        }
+    };
+    return {.natural_endpoint = return_host_id_if_not_left(paired_replica),
+            .endpoint_with_no_pairing = return_host_id_if_not_left(no_pairing_replica)};
 }
 
 static future<> apply_to_remote_endpoints(service::storage_proxy& proxy, locator::effective_replication_map_ptr ermp,
@@ -2041,6 +2136,12 @@ future<> view_update_generator::mutate_MV(
 {
     auto& ks = _db.find_keyspace(base->ks_name());
     auto& replication = ks.get_replication_strategy();
+    // We set legacy self-pairing for old vnode-based tables (for backward
+    // compatibility), and unset it for tablets - where range movements
+    // are more frequent and backward compatibility is less important.
+    // TODO: Maybe allow users to set use_legacy_self_pairing explicitly
+    // on a view, like we have the synchronous_updates_flag.
+    bool use_legacy_self_pairing = !ks.uses_tablets();
     std::unordered_map<table_id, locator::effective_replication_map_ptr> erms;
     auto get_erm = [&] (table_id id) {
         auto it = erms.find(id);
@@ -2053,6 +2154,10 @@ future<> view_update_generator::mutate_MV(
     for (const auto& mut : view_updates) {
         (void)get_erm(mut.s->id());
     }
+    // Enable rack-aware view updates pairing for tablets
+    // when the cluster feature is enabled so that all replicas agree
+    // on the pairing algorithm.
+    bool use_tablets_rack_aware_view_pairing = _db.features().tablet_rack_aware_view_pairing && ks.uses_tablets();
     auto me = base_ermp->get_topology().my_host_id();
     static constexpr size_t max_concurrent_updates = 128;
     co_await utils::get_local_injector().inject("delay_before_get_view_natural_endpoint", 8000ms);
@@ -2060,7 +2165,7 @@ future<> view_update_generator::mutate_MV(
         auto view_token = dht::get_token(*mut.s, mut.fm.key());
         auto view_ermp = erms.at(mut.s->id());
         auto [target_endpoint, no_pairing_endpoint] = get_view_natural_endpoint(me, base_ermp, view_ermp, replication, base_token, view_token,
-                ks.uses_tablets(), cf_stats);
+                use_legacy_self_pairing, use_tablets_rack_aware_view_pairing, cf_stats);
         auto remote_endpoints = view_ermp->get_pending_replicas(view_token);
         auto memory_units = seastar::make_lw_shared<db::timeout_semaphore_units>(pending_view_update_memory_units.split(memory_usage_of(mut)));
         if (no_pairing_endpoint) {
@@ -3492,7 +3597,7 @@ view_updating_consumer::view_updating_consumer(view_update_generator& gen, schem
     })
 { }
 
-delete_ghost_rows_visitor::delete_ghost_rows_visitor(service::storage_proxy& proxy, service::query_state& state, view_ptr view, db::timeout_clock::duration timeout_duration, size_t concurrency, std::exception_ptr& ex)
+delete_ghost_rows_visitor::delete_ghost_rows_visitor(service::storage_proxy& proxy, service::query_state& state, view_ptr view, db::timeout_clock::duration timeout_duration)
         : _proxy(proxy)
         , _state(state)
         , _timeout_duration(timeout_duration)
@@ -3500,20 +3605,8 @@ delete_ghost_rows_visitor::delete_ghost_rows_visitor(service::storage_proxy& pro
         , _view_table(_proxy.get_db().local().find_column_family(view))
         , _base_schema(_proxy.get_db().local().find_schema(_view->view_info()->base_id()))
         , _view_pk()
-        , _concurrency_semaphore(concurrency)
-        , _ex(ex)
 {}
 
-
-delete_ghost_rows_visitor::~delete_ghost_rows_visitor() noexcept {
-    try {
-        _gate.close().get();
-    } catch (...) {
-        // Closing the gate should never throw, but if it does anyway, capture the exception.
-        _ex = std::current_exception();
-    }
-}
-
 void delete_ghost_rows_visitor::accept_new_partition(const partition_key& key, uint32_t row_count) {
     SCYLLA_ASSERT(thread::running_in_thread());
     _view_pk = key;
@@ -3521,18 +3614,7 @@ void delete_ghost_rows_visitor::accept_new_partition(const partition_key& key, u
 
 // Assumes running in seastar::thread
 void delete_ghost_rows_visitor::accept_new_row(const clustering_key& ck, const query::result_row_view& static_row, const query::result_row_view& row) {
-    auto units = get_units(_concurrency_semaphore, 1).get();
-    (void)seastar::try_with_gate(_gate, [this, pk = _view_pk.value(), units = std::move(units), ck] () mutable {
-        return do_accept_new_row(std::move(pk), std::move(ck)).then_wrapped([this, units = std::move(units)] (future<>&& f) mutable {
-            if (f.failed()) {
-                _ex = f.get_exception();
-            }
-        });
-    });
-}
-
-future<> delete_ghost_rows_visitor::do_accept_new_row(partition_key pk, clustering_key ck) {
-    auto view_exploded_pk = pk.explode();
+    auto view_exploded_pk = _view_pk->explode();
     auto view_exploded_ck = ck.explode();
     std::vector<bytes> base_exploded_pk(_base_schema->partition_key_size());
     std::vector<bytes> base_exploded_ck(_base_schema->clustering_key_size());
@@ -3567,17 +3649,17 @@ future<> delete_ghost_rows_visitor::do_accept_new_row(partition_key pk, clusteri
             _proxy.get_max_result_size(partition_slice), query::tombstone_limit(_proxy.get_tombstone_limit()));
     auto timeout = db::timeout_clock::now() + _timeout_duration;
     service::storage_proxy::coordinator_query_options opts{timeout, _state.get_permit(), _state.get_client_state(), _state.get_trace_state()};
-    auto base_qr = co_await _proxy.query(_base_schema, command, std::move(partition_ranges), db::consistency_level::ALL, opts);
+    auto base_qr = _proxy.query(_base_schema, command, std::move(partition_ranges), db::consistency_level::ALL, opts).get();
     query::result& result = *base_qr.query_result;
-    auto delete_ghost_row = [&]() -> future<> {
-        mutation m(_view, pk);
+    auto delete_ghost_row = [&]() {
+        mutation m(_view, *_view_pk);
         auto& row = m.partition().clustered_row(*_view, ck);
         row.apply(tombstone(api::new_timestamp(), gc_clock::now()));
         timeout = db::timeout_clock::now() + _timeout_duration;
-        return _proxy.mutate({m}, db::consistency_level::ALL, timeout, _state.get_trace_state(), empty_service_permit(), db::allow_per_partition_rate_limit::no);
+        _proxy.mutate({m}, db::consistency_level::ALL, timeout, _state.get_trace_state(), empty_service_permit(), db::allow_per_partition_rate_limit::no).get();
     };
     if (result.row_count().value_or(0) == 0) {
-        co_await delete_ghost_row();
+        delete_ghost_row();
     } else if (!view_key_cols_not_in_base_key.empty()) {
         if (result.row_count().value_or(0) != 1) {
             on_internal_error(vlogger, format("Got multiple base rows corresponding to a single view row when pruning {}.{}", _view->ks_name(), _view->cf_name()));
@@ -3587,7 +3669,7 @@ future<> delete_ghost_rows_visitor::do_accept_new_row(partition_key pk, clusteri
         for (const auto& [col_def, col_val] : view_key_cols_not_in_base_key) {
             const data_value* base_val = base_row.get_data_value(col_def->name_as_text());
             if (!base_val || base_val->is_null() || col_val != base_val->serialize_nonnull()) {
-                co_await delete_ghost_row();
+                delete_ghost_row();
                 break;
             }
         }

db/view/view.hh

@@ -305,7 +305,8 @@ endpoints_to_update get_view_natural_endpoint(
     const locator::abstract_replication_strategy& replication_strategy,
     const dht::token& base_token,
     const dht::token& view_token,
-    bool use_tablets,
+    bool use_legacy_self_pairing,
+    bool use_tablets_basic_rack_aware_view_pairing,
     replica::cf_stats& cf_stats);
 
 /// Verify that the provided keyspace is eligible for storing materialized views.

db/view/view_building_worker.cc

@@ -198,7 +198,6 @@ future<> view_building_worker::register_staging_sstable_tasks(std::vector<sstabl
 
 future<> view_building_worker::run_staging_sstables_registrator() {
     while (!_as.abort_requested()) {
-        bool sleep = false;
         try {
             auto lock = co_await get_units(_staging_sstables_mutex, 1, _as);
             co_await create_staging_sstable_tasks();
@@ -215,14 +214,6 @@ future<> view_building_worker::run_staging_sstables_registrator() {
             vbw_logger.warn("Got group0_concurrent_modification while creating staging sstable tasks");
         } catch (raft::request_aborted&) {
             vbw_logger.warn("Got raft::request_aborted while creating staging sstable tasks");
-        } catch (...) {
-            vbw_logger.error("Exception while creating staging sstable tasks: {}", std::current_exception());
-            sleep = true;
-        }
-
-        if (sleep) {
-            vbw_logger.debug("Sleeping after exception.");
-            co_await seastar::sleep_abortable(1s, _as).handle_exception([] (auto x) { return make_ready_future<>(); });
         }
     }
 }
@@ -426,12 +417,9 @@ future<> view_building_worker::check_for_aborted_tasks() {
 
         auto my_host_id = vbw._db.get_token_metadata().get_topology().my_host_id();
         auto my_replica = locator::tablet_replica{my_host_id, this_shard_id()};
-        auto it = vbw._state._batch->tasks.begin();
-        while (it != vbw._state._batch->tasks.end()) {
-            auto id = it->first;
-            auto task_opt = building_state.get_task(it->second.base_id, my_replica, id);
-
-            ++it; // Advance the iterator before potentially removing the entry from the map.
+        auto tasks_map = vbw._state._batch->tasks; // Potentially, we'll remove elements from the map, so we need a copy to iterate over it
+        for (auto& [id, t]: tasks_map) {
+            auto task_opt = building_state.get_task(t.base_id, my_replica, id);
             if (!task_opt || task_opt->get().aborted) {
                 co_await vbw._state._batch->abort_task(id);
             }
@@ -461,7 +449,7 @@ static std::unordered_set<table_id> get_ids_of_all_views(replica::database& db,
     }) | std::ranges::to<std::unordered_set>();;
 }
 
-// If `state::processing_base_table` is different that the `view_building_state::currently_processed_base_table`,
+// If `state::processing_base_table` is diffrent that the `view_building_state::currently_processed_base_table`,
 // clear the state, save and flush new base table
 future<> view_building_worker::state::update_processing_base_table(replica::database& db, const view_building_state& building_state, abort_source& as) {
     if (processing_base_table != building_state.currently_processed_base_table) {
@@ -583,6 +571,8 @@ future<> view_building_worker::batch::do_work() {
             break;
         }
     }
+
+    _vbw.local()._vb_state_machine.event.broadcast();
 }
 
 future<> view_building_worker::do_build_range(table_id base_id, std::vector<table_id> views_ids, dht::token last_token, abort_source& as) {
@@ -784,15 +774,13 @@ future<std::vector<utils::UUID>> view_building_worker::work_on_tasks(raft::term_
             tasks.insert({id, *task_opt});
         }
 #ifdef SEASTAR_DEBUG
-        {
-            auto& some_task = tasks.begin()->second;
-            for (auto& [_, t]: tasks) {
-                SCYLLA_ASSERT(t.base_id == some_task.base_id);
-                SCYLLA_ASSERT(t.last_token == some_task.last_token);
-                SCYLLA_ASSERT(t.replica == some_task.replica);
-                SCYLLA_ASSERT(t.type == some_task.type);
-                SCYLLA_ASSERT(t.replica.shard == this_shard_id());
-            }
+        auto& some_task = tasks.begin()->second;
+        for (auto& [_, t]: tasks) {
+            SCYLLA_ASSERT(t.base_id == some_task.base_id);
+            SCYLLA_ASSERT(t.last_token == some_task.last_token);
+            SCYLLA_ASSERT(t.replica == some_task.replica);
+            SCYLLA_ASSERT(t.type == some_task.type);
+            SCYLLA_ASSERT(t.replica.shard == this_shard_id());
         }
 #endif
 
@@ -823,6 +811,25 @@ future<std::vector<utils::UUID>> view_building_worker::work_on_tasks(raft::term_
     co_return collect_completed_tasks();
 }
 
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
 }
 
 }

db/virtual_tables.cc

@@ -749,7 +749,6 @@ class clients_table : public streaming_virtual_table {
             .with_column("ssl_protocol", utf8_type)
             .with_column("username", utf8_type)
             .with_column("scheduling_group", utf8_type)
-            .with_column("client_options", map_type_impl::get_instance(utf8_type, utf8_type, false))
             .with_hash_version()
             .build();
     }
@@ -767,7 +766,7 @@ class clients_table : public streaming_virtual_table {
 
     future<> execute(reader_permit permit, result_collector& result, const query_restrictions& qr) override {
         // Collect
-        using client_data_vec = utils::chunked_vector<foreign_ptr<std::unique_ptr<client_data>>>;
+        using client_data_vec = utils::chunked_vector<client_data>;
         using shard_client_data = std::vector<client_data_vec>;
         std::vector<foreign_ptr<std::unique_ptr<shard_client_data>>> cd_vec;
         cd_vec.resize(smp::count);
@@ -807,13 +806,13 @@ class clients_table : public streaming_virtual_table {
         for (unsigned i = 0; i < smp::count; i++) {
             for (auto&& ps_cdc : *cd_vec[i]) {
                 for (auto&& cd : ps_cdc) {
-                    if (cd_map.contains(cd->ip)) {
-                        cd_map[cd->ip].emplace_back(std::move(cd));
+                    if (cd_map.contains(cd.ip)) {
+                        cd_map[cd.ip].emplace_back(std::move(cd));
                     } else {
-                        dht::decorated_key key = make_partition_key(cd->ip);
+                        dht::decorated_key key = make_partition_key(cd.ip);
                         if (this_shard_owns(key) && contains_key(qr.partition_range(), key)) {
-                            ips.insert(decorated_ip{std::move(key), cd->ip});
-                            cd_map[cd->ip].emplace_back(std::move(cd));
+                            ips.insert(decorated_ip{std::move(key), cd.ip});
+                            cd_map[cd.ip].emplace_back(std::move(cd));
                         }
                     }
                     co_await coroutine::maybe_yield();
@@ -826,58 +825,39 @@ class clients_table : public streaming_virtual_table {
             co_await result.emit_partition_start(dip.key);
             auto& clients = cd_map[dip.ip];
 
-            std::ranges::sort(clients, [] (const foreign_ptr<std::unique_ptr<client_data>>& a, const foreign_ptr<std::unique_ptr<client_data>>& b) {
-                return a->port < b->port || a->client_type_str() < b->client_type_str();
+            std::ranges::sort(clients, [] (const client_data& a, const client_data& b) {
+                return a.port < b.port || a.client_type_str() < b.client_type_str();
             });
 
             for (const auto& cd : clients) {
-                clustering_row cr(make_clustering_key(cd->port, cd->client_type_str()));
-                set_cell(cr.cells(), "shard_id", cd->shard_id);
-                set_cell(cr.cells(), "connection_stage", cd->stage_str());
-                if (cd->driver_name) {
-                    set_cell(cr.cells(), "driver_name", cd->driver_name->key());
+                clustering_row cr(make_clustering_key(cd.port, cd.client_type_str()));
+                set_cell(cr.cells(), "shard_id", cd.shard_id);
+                set_cell(cr.cells(), "connection_stage", cd.stage_str());
+                if (cd.driver_name) {
+                    set_cell(cr.cells(), "driver_name", *cd.driver_name);
                 }
-                if (cd->driver_version) {
-                    set_cell(cr.cells(), "driver_version", cd->driver_version->key());
+                if (cd.driver_version) {
+                    set_cell(cr.cells(), "driver_version", *cd.driver_version);
                 }
-                if (cd->hostname) {
-                    set_cell(cr.cells(), "hostname", *cd->hostname);
+                if (cd.hostname) {
+                    set_cell(cr.cells(), "hostname", *cd.hostname);
                 }
-                if (cd->protocol_version) {
-                    set_cell(cr.cells(), "protocol_version", *cd->protocol_version);
+                if (cd.protocol_version) {
+                    set_cell(cr.cells(), "protocol_version", *cd.protocol_version);
                 }
-                if (cd->ssl_cipher_suite) {
-                    set_cell(cr.cells(), "ssl_cipher_suite", *cd->ssl_cipher_suite);
+                if (cd.ssl_cipher_suite) {
+                    set_cell(cr.cells(), "ssl_cipher_suite", *cd.ssl_cipher_suite);
                 }
-                if (cd->ssl_enabled) {
-                    set_cell(cr.cells(), "ssl_enabled", *cd->ssl_enabled);
+                if (cd.ssl_enabled) {
+                    set_cell(cr.cells(), "ssl_enabled", *cd.ssl_enabled);
                 }
-                if (cd->ssl_protocol) {
-                    set_cell(cr.cells(), "ssl_protocol", *cd->ssl_protocol);
+                if (cd.ssl_protocol) {
+                    set_cell(cr.cells(), "ssl_protocol", *cd.ssl_protocol);
                 }
-                set_cell(cr.cells(), "username", cd->username ? *cd->username : sstring("anonymous"));
-                if (cd->scheduling_group_name) {
-                    set_cell(cr.cells(), "scheduling_group", *cd->scheduling_group_name);
+                set_cell(cr.cells(), "username", cd.username ? *cd.username : sstring("anonymous"));
+                if (cd.scheduling_group_name) {
+                    set_cell(cr.cells(), "scheduling_group", *cd.scheduling_group_name);
                 }
-
-                auto map_type = map_type_impl::get_instance(
-                    utf8_type,
-                    utf8_type,
-                    false
-                );
-
-                auto prepare_client_options = [] (const auto& client_options) {
-                    map_type_impl::native_type tmp;
-                    for (auto& co: client_options) {
-                        auto map_element = std::make_pair(data_value(co.key.key()), data_value(co.value.key()));
-                        tmp.push_back(std::move(map_element));
-                    }
-                    return tmp;
-                };
-
-                set_cell(cr.cells(), "client_options",
-                    make_map_value(map_type, prepare_client_options(cd->client_options)));
-
                 co_await result.emit_row(std::move(cr));
             }
             co_await result.emit_partition_end();

dist/common/sysconfig/scylla-node-exporter

@@ -1 +1 @@
-SCYLLA_NODE_EXPORTER_ARGS="--collector.interrupts --collector.ethtool.metrics-include='(bw_in_allowance_exceeded|bw_out_allowance_exceeded|conntrack_allowance_exceeded|conntrack_allowance_available|linklocal_allowance_exceeded)' --collector.ethtool --no-collector.hwmon --no-collector.bcache --no-collector.btrfs --no-collector.fibrechannel --no-collector.infiniband --no-collector.ipvs --no-collector.nfs --no-collector.nfsd --no-collector.powersupplyclass --no-collector.rapl --no-collector.tapestats --no-collector.thermal_zone --no-collector.udp_queues --no-collector.zfs"
+SCYLLA_NODE_EXPORTER_ARGS="--collector.interrupts --no-collector.hwmon --no-collector.bcache --no-collector.btrfs --no-collector.fibrechannel --no-collector.infiniband --no-collector.ipvs --no-collector.nfs --no-collector.nfsd --no-collector.powersupplyclass --no-collector.rapl --no-collector.tapestats --no-collector.thermal_zone --no-collector.udp_queues --no-collector.zfs"

dist/debian/debian/scylla-server.install

@@ -2,6 +2,7 @@ etc/default/scylla-server
 etc/default/scylla-housekeeping
 etc/scylla.d/*.conf
 etc/bash_completion.d/nodetool-completion
+opt/scylladb/share/p11-kit/modules/*
 opt/scylladb/share/doc/scylla/*
 opt/scylladb/share/doc/scylla/licenses/
 usr/lib/systemd/system/*.timer

dist/redhat/scylla.spec

@@ -122,6 +122,7 @@ ln -sfT /etc/scylla /var/lib/scylla/conf
 %config(noreplace) %{_sysconfdir}/sysconfig/scylla-housekeeping
 %attr(0755,root,root) %dir %{_sysconfdir}/scylla.d
 %config(noreplace) %{_sysconfdir}/scylla.d/*.conf
+/opt/scylladb/share/p11-kit/modules/*
 /opt/scylladb/share/doc/scylla/*
 %{_unitdir}/scylla-fstrim.service
 %{_unitdir}/scylla-housekeeping-daily.service

docs/README.md

@@ -71,7 +71,7 @@ Use "Bash on Ubuntu on Windows" for the same tools and capabilities as on Linux
 
 ### Building the Docs 
 
-1. Run `make preview` in the `docs/` directory to build the documentation.
+1. Run `make preview` to build the documentation.
 1. Preview the built documentation locally at http://127.0.0.1:5500/.
 
 ### Cleanup

docs/_ext/scylladb_metrics.py

@@ -41,8 +41,6 @@ class MetricsProcessor:
                 # Get metrics from the file
                 try:
                     metrics_file = metrics.get_metrics_from_file(relative_path, "scylla_", metrics_info, strict=strict)
-                except SystemExit:
-                    pass
                 finally:
                     os.chdir(old_cwd)
                 if metrics_file:

docs/_utils/redirects.yaml

@@ -1,18 +1,6 @@
 ### a dictionary of redirections
 #old path: new path
 
-# Move the diver information to another project
-
-/stable/using-scylla/drivers/index.html: https://docs.scylladb.com/stable/drivers/index.html
-/stable/using-scylla/drivers/dynamo-drivers/index.html: https://docs.scylladb.com/stable/drivers/dynamo-drivers.html
-/stable/using-scylla/drivers/cql-drivers/index.html: https://docs.scylladb.com/stable/drivers/cql-drivers.html
-/stable/using-scylla/drivers/cql-drivers/scylla-python-driver.html: https://docs.scylladb.com/stable/drivers/cql-drivers.html
-/stable/using-scylla/drivers/cql-drivers/scylla-java-driver.html: https://docs.scylladb.com/stable/drivers/cql-drivers.html
-/stable/using-scylla/drivers/cql-drivers/scylla-go-driver.html: https://docs.scylladb.com/stable/drivers/cql-drivers.html
-/stable/using-scylla/drivers/cql-drivers/scylla-gocqlx-driver.html: https://docs.scylladb.com/stable/drivers/cql-drivers.html
-/stable/using-scylla/drivers/cql-drivers/scylla-cpp-driver.html: https://docs.scylladb.com/stable/drivers/cql-drivers.html
-/stable/using-scylla/drivers/cql-drivers/scylla-rust-driver.html: https://docs.scylladb.com/stable/drivers/cql-drivers.html
-
 # Redirect 2025.1 upgrade guides that are not on master but were indexed by Google (404 reported)
 
 /master/upgrade/upgrade-guides/upgrade-guide-from-2024.x-to-2025.1/upgrade-guide-from-2024.x-to-2025.1.html: https://docs.scylladb.com/manual/stable/upgrade/index.html

docs/alternator/alternator.md

@@ -1,17 +1,17 @@
-# Alternator: DynamoDB API in ScyllaDB
+# Alternator: DynamoDB API in Scylla
 
 ## Introduction
-Alternator is a ScyllaDB feature adding compatibility with Amazon DynamoDB(TM).
+Alternator is a Scylla feature adding compatibility with Amazon DynamoDB(TM).
 DynamoDB's API uses JSON-encoded requests and responses which are sent over
 an HTTP or HTTPS transport. It is described in detail in Amazon's [DynamoDB
 API Reference](https://docs.aws.amazon.com/amazondynamodb/latest/APIReference/).
 
 Our goal is that any application written to use Amazon DynamoDB could
-be run, unmodified, against ScyllaDB with Alternator enabled. Alternator's
+be run, unmodified, against Scylla with Alternator enabled. Alternator's
 compatibility with DynamoDB is fairly complete, but users should be aware
 of some differences and some unimplemented features. The extent of
 Alternator's compatibility with DynamoDB is described in the
-[ScyllaDB Alternator for DynamoDB users](compatibility.md) document,
+[Scylla Alternator for DynamoDB users](compatibility.md) document,
 which is updated as the work on Alternator progresses and compatibility
 continues to improve.
 
@@ -19,8 +19,8 @@ Alternator also adds several features and APIs that are not available in
 DynamoDB. These are described in [Alternator-specific APIs](new-apis.md).
 
 ## Running Alternator
-By default, ScyllaDB does not listen for DynamoDB API requests. To enable
-this API in ScyllaDB you must set at least two configuration options,
+By default, Scylla does not listen for DynamoDB API requests. To enable
+this API in Scylla you must set at least two configuration options,
 **alternator_port** and **alternator_write_isolation**. For example in the
 YAML configuration file:
 ```yaml
@@ -30,7 +30,7 @@ alternator_write_isolation: only_rmw_uses_lwt # or always, forbid or unsafe
 or, equivalently, via command-line arguments: `--alternator-port=8000
 --alternator-write-isolation=only_rmw_uses_lwt.
 
-the **alternator_port** option determines on which port ScyllaDB listens for
+the **alternator_port** option determines on which port Scylla listens for
 DynamoDB API requests. By default, it listens on this port on all network
 interfaces. To listen only on a specific interface, configure also the
 **alternator_address** option.
@@ -41,12 +41,12 @@ Alternator has four different choices
 for the implementation of writes, each with different advantages. You should
 carefully consider which of the options makes more sense for your intended
 use case and configure alternator_write_isolation accordingly. There is
-currently no default for this option: Trying to run ScyllaDB with an Alternator
+currently no default for this option: Trying to run Scylla with an Alternator
 port selected but without configuring write isolation will result in an error message,
 asking you to set it.
 
 In addition to (or instead of) serving HTTP requests on alternator_port,
-ScyllaDB can accept DynamoDB API requests over HTTPS (encrypted), on the port
+Scylla can accept DynamoDB API requests over HTTPS (encrypted), on the port
 specified by **alternator_https_port**. As usual for HTTPS servers, the
 operator must specify certificate and key files. By default these should
 be placed in `/etc/scylla/scylla.crt` and `/etc/scylla/scylla.key`, but
@@ -54,7 +54,7 @@ these default locations can overridden by specifying
 `--alternator-encryption-options keyfile="..."` and
 `--alternator-encryption-options certificate="..."`.
 
-By default, ScyllaDB saves a snapshot of deleted tables. But Alternator does
+By default, Scylla saves a snapshot of deleted tables. But Alternator does
 not offer an API to restore these snapshots, so these snapshots are not useful
 and waste disk space - deleting a table does not recover any disk space.
 It is therefore recommended to disable this automatic-snapshotting feature
@@ -73,11 +73,11 @@ itself. Instructions, code and examples for doing this can be found in the
 
 This section provides only a very brief introduction to Alternator's
 design. A much more detailed document about the features of the DynamoDB
-API and how they are, or could be, implemented in ScyllaDB can be found in:
+API and how they are, or could be, implemented in Scylla can be found in:
 <https://docs.google.com/document/d/1i4yjF5OSAazAY_-T8CBce9-2ykW4twx_E_Nt2zDoOVs>
 
 Almost all of Alternator's source code (except some initialization code)
-can be found in the alternator/ subdirectory of ScyllaDB's source code.
+can be found in the alternator/ subdirectory of Scylla's source code.
 Extensive functional tests can be found in the test/alternator
 subdirectory. These tests are written in Python, and can be run against
 both Alternator and Amazon's DynamoDB; This allows verifying that
@@ -85,15 +85,15 @@ Alternator's behavior matches the one observed on DynamoDB.
 See test/alternator/README.md for more information about the tests and
 how to run them.
 
-With Alternator enabled on port 8000 (for example), every ScyllaDB node
+With Alternator enabled on port 8000 (for example), every Scylla node
 listens for DynamoDB API requests on this port. These requests, in
 JSON format over HTTP, are parsed and result in calls to internal Scylla
 C++ functions - there is no CQL generation or parsing involved.
-In ScyllaDB terminology, the node receiving the request acts as the
+In Scylla terminology, the node receiving the request acts as the
 *coordinator*, and often passes the request on to one or more other nodes -
 *replicas* which hold copies of the requested data.
 
-Alternator tables are stored as ScyllaDB tables, each in a separate keyspace.
+Alternator tables are stored as Scylla tables, each in a separate keyspace.
 Each keyspace is initialized when the corresponding Alternator table is
 created (with a CreateTable request). The replication factor (RF) for this
 keyspace is chosen at that point, depending on the size of the cluster:
@@ -101,19 +101,19 @@ RF=3 is used on clusters with three or more nodes, and RF=1 is used for
 smaller clusters. Such smaller clusters are, of course, only recommended
 for tests because of the risk of data loss.
 
-Each table in Alternator is stored as a ScyllaDB table in a separate
+Each table in Alternator is stored as a Scylla table in a separate
 keyspace. The DynamoDB key columns (hash and sort key) have known types,
-and become partition and clustering key columns of the ScyllaDB table.
+and become partition and clustering key columns of the Scylla table.
 All other attributes may be different for each row, so are stored in one
-map column in ScyllaDB, and not as separate columns.
+map column in Scylla, and not as separate columns.
 
 DynamoDB supports two consistency levels for reads, "eventual consistency"
-and "strong consistency". These two modes are implemented using ScyllaDB's CL
+and "strong consistency". These two modes are implemented using Scylla's CL
 (consistency level) feature: All writes are done using the `LOCAL_QUORUM`
 consistency level, then strongly-consistent reads are done with
 `LOCAL_QUORUM`, while eventually-consistent reads are with just `LOCAL_ONE`.
 
-In ScyllaDB (and its inspiration, Cassandra), high write performance is
+In Scylla (and its inspiration, Cassandra), high write performance is
 achieved by ensuring that writes do not require reads from disk.
 The DynamoDB API, however, provides many types of requests that need a read
 before the write (a.k.a. RMW requests - read-modify-write). For example,
@@ -121,7 +121,7 @@ a request may copy an existing attribute, increment an attribute,
 be conditional on some expression involving existing values of attribute,
 or request that the previous values of attributes be returned. These
 read-modify-write transactions should be _isolated_ from each other, so
-by default Alternator implements every write operation using ScyllaDB's
+by default Alternator implements every write operation using Scylla's
 LWT (lightweight transactions). This default can be overridden on a per-table
 basis, by tagging the table as explained above in the "write isolation
 policies" section.

docs/alternator/compatibility.md

@@ -1,6 +1,6 @@
 # ScyllaDB Alternator for DynamoDB users
 
-ScyllaDB supports the DynamoDB API (this feature is codenamed "Alternator").
+Scylla supports the DynamoDB API (this feature is codenamed "Alternator").
 Our goal is to support any application written for Amazon DynamoDB.
 Nevertheless, there are a few differences between DynamoDB and Scylla, and
 and a few DynamoDB features that have not yet been implemented in Scylla.
@@ -8,16 +8,16 @@ The purpose of this document is to inform users of these differences.
 
 ## Provisioning
 
-The most obvious difference between DynamoDB and ScyllaDB is that while
-DynamoDB is a shared cloud service, ScyllaDB is a dedicated service running
+The most obvious difference between DynamoDB and Scylla is that while
+DynamoDB is a shared cloud service, Scylla is a dedicated service running
 on your private cluster. Whereas DynamoDB allows you to "provision" the
 number of requests per second you'll need - or at an extra cost not even
-provision that - ScyllaDB requires you to provision your cluster. You need
+provision that - Scylla requires you to provision your cluster. You need
 to reason about the number and size of your nodes - not the throughput.
 
 Moreover, DynamoDB's per-table provisioning (`BillingMode=PROVISIONED`) is
 not yet supported by Scylla. The BillingMode and ProvisionedThroughput options
-on a table need to be valid but are ignored, and ScyllaDB behaves like DynamoDB's
+on a table need to be valid but are ignored, and Scylla behaves like DynamoDB's
 `BillingMode=PAY_PER_REQUEST`: All requests are accepted without a per-table
 throughput cap.
 
@@ -33,7 +33,7 @@ Instructions for doing this can be found in:
 
 ## Write isolation policies
 
-ScyllaDB was designed to optimize the performance of pure write operations -
+Scylla was designed to optimize the performance of pure write operations -
 writes which do not need to read the previous value of the item.
 In CQL, writes which do need the previous value of the item must explicitly
 use the slower LWT ("LightWeight Transaction") feature to be correctly
@@ -79,11 +79,11 @@ a _higher_ timestamp - and this will be the "last write" that wins.
 To avoid or mitigate this write reordering issue, users may consider
 one or more of the following:
 
-1. Use NTP to keep the clocks on the different ScyllaDB nodes synchronized.
+1. Use NTP to keep the clocks on the different Scylla nodes synchronized.
    If the delay between the two writes is longer than NTP's accuracy,
    they will not be reordered.
 2. If an application wants to ensure that two specific writes are not
-   reordered, it should send both requests to the same ScyllaDB node.
+   reordered, it should send both requests to the same Scylla node.
    Care should be taken when using a load balancer - which might redirect
    two requests to two different nodes.
 3. Consider using the `always_use_lwt` write isolation policy.
@@ -210,7 +210,7 @@ CREATE SERVICE_LEVEL IF NOT EXISTS oltp WITH SHARES = 1000;
 ATTACH SERVICE_LEVEL olap TO alice;
 ATTACH SERVICE_LEVEL oltp TO bob;
 ```
-Note that `alternator_enforce_authorization` has to be enabled in ScyllaDB configuration.
+Note that `alternator_enforce_authorization` has to be enabled in Scylla configuration.
 
 See [Authorization](##Authorization) section to learn more about roles and authorization.
 See [Workload Prioritization](../features/workload-prioritization)
@@ -218,11 +218,11 @@ to read about Workload Prioritization in detail.
 
 ## Metrics
 
-ScyllaDB has an advanced and extensive monitoring framework for inspecting
-and graphing hundreds of different metrics of ScyllaDB's usage and performance.
-ScyllaDB's monitoring stack, based on Grafana and Prometheus, is described in
+Scylla has an advanced and extensive monitoring framework for inspecting
+and graphing hundreds of different metrics of Scylla's usage and performance.
+Scylla's monitoring stack, based on Grafana and Prometheus, is described in
 <https://docs.scylladb.com/operating-scylla/monitoring/>.
-This monitoring stack is different from DynamoDB's offering - but ScyllaDB's
+This monitoring stack is different from DynamoDB's offering - but Scylla's
 is significantly more powerful and gives the user better insights on
 the internals of the database and its performance.
 
@@ -248,7 +248,7 @@ data in different partition order. Applications mustn't rely on that
 undocumented order.
 
 Note that inside each partition, the individual items will be sorted the same
-in DynamoDB and ScyllaDB - determined by the _sort key_ defined for that table.
+in DynamoDB and Scylla - determined by the _sort key_ defined for that table.
 
 ---
 
@@ -274,7 +274,7 @@ is different, or can be configured in Alternator:
 ## Experimental API features
 
 Some DynamoDB API features are supported by Alternator, but considered
-**experimental** in this release. An experimental feature in ScyllaDB is a
+**experimental** in this release. An experimental feature in Scylla is a
 feature whose functionality is complete, or mostly complete, but it is not
 as thoroughly tested or optimized as regular features. Also, an experimental
 feature's implementation is still subject to change and upgrades may not be
@@ -351,8 +351,8 @@ they should be easy to detect. Here is a list of these unimplemented features:
 
 * The on-demand backup APIs are not supported: CreateBackup, DescribeBackup,
   DeleteBackup, ListBackups, RestoreTableFromBackup.
-  For now, users can use ScyllaDB's existing backup solutions such as snapshots
-  or ScyllaDB Manager.
+  For now, users can use Scylla's existing backup solutions such as snapshots
+  or Scylla Manager.
   <https://github.com/scylladb/scylla/issues/5063>
 
 * Continuous backup (the ability to restore any point in time) is also not
@@ -370,7 +370,7 @@ they should be easy to detect. Here is a list of these unimplemented features:
   <https://github.com/scylladb/scylla/issues/5068>
 
 * DAX (DynamoDB Accelerator), an in-memory cache for DynamoDB, is not
-  available in for Alternator. Anyway, it should not be necessary - ScyllaDB's
+  available in for Alternator. Anyway, it should not be necessary - Scylla's
   internal cache is already rather advanced and there is no need to place
   another cache in front of the it. We wrote more about this here:
   <https://www.scylladb.com/2017/07/31/database-caches-not-good/>
@@ -384,7 +384,7 @@ they should be easy to detect. Here is a list of these unimplemented features:
 * The PartiQL syntax (SQL-like SELECT/UPDATE/INSERT/DELETE expressions)
   and the operations ExecuteStatement, BatchExecuteStatement and
   ExecuteTransaction are not yet supported.
-  A user that is interested in an SQL-like syntax can consider using ScyllaDB's
+  A user that is interested in an SQL-like syntax can consider using Scylla's
   CQL protocol instead.
   This feature was added to DynamoDB in November 2020.
   <https://github.com/scylladb/scylla/issues/8787>
@@ -393,7 +393,7 @@ they should be easy to detect. Here is a list of these unimplemented features:
   which is different from AWS's. In particular, the operations
   DescribeContributorInsights, ListContributorInsights and
   UpdateContributorInsights that configure Amazon's "CloudWatch Contributor
-  Insights" are not yet supported. ScyllaDB has different ways to retrieve the
+  Insights" are not yet supported. Scylla has different ways to retrieve the
   same information, such as which items were accessed most often.
   <https://github.com/scylladb/scylla/issues/8788>
 

docs/alternator/getting-started.md

@@ -11,7 +11,7 @@ This section will guide you through the steps for setting up the cluster:
    <https://hub.docker.com/r/scylladb/scylla/>, but add to every `docker run`
    command a `-p 8000:8000` before the image name and
    `--alternator-port=8000 --alternator-write-isolation=always` at the end.
-   The "alternator-port" option specifies on which port ScyllaDB will listen for
+   The "alternator-port" option specifies on which port Scylla will listen for
    the (unencrypted) DynamoDB API, and the "alternator-write-isolation" chooses
    whether or not Alternator will use LWT for every write.
    For example,
@@ -24,10 +24,10 @@ This section will guide you through the steps for setting up the cluster:
 By default, ScyllaDB run in this way will not have authentication or
 authorization enabled, and any DynamoDB API request will be honored without
 requiring them to be signed appropriately. See the
-[ScyllaDB Alternator for DynamoDB users](compatibility.md#authentication-and-authorization)
+[Scylla Alternator for DynamoDB users](compatibility.md#authentication-and-authorization)
 document on how to configure authentication and authorization.
 
-## Testing ScyllaDB's DynamoDB API support:
+## Testing Scylla's DynamoDB API support:
 ### Running AWS Tic Tac Toe demo app to test the cluster:
 1. Follow the instructions on the [AWS github page](https://github.com/awsdocs/amazon-dynamodb-developer-guide/blob/master/doc_source/TicTacToe.Phase1.md)
 2. Enjoy your tic-tac-toe game :-)

docs/alternator/new-apis.md

@@ -2,9 +2,9 @@
 
 Alternator's primary goal is to be compatible with Amazon DynamoDB(TM)
 and its APIs, so that any application written to use Amazon DynamoDB could
-be run, unmodified, against ScyllaDB with Alternator enabled. The extent of
+be run, unmodified, against Scylla with Alternator enabled. The extent of
 Alternator's compatibility with DynamoDB is described in the
-[ScyllaDB Alternator for DynamoDB users](compatibility.md) document.
+[Scylla Alternator for DynamoDB users](compatibility.md) document.
 
 But Alternator also adds several features and APIs that are not available in
 DynamoDB. These Alternator-specific APIs are documented here.
@@ -15,7 +15,7 @@ _conditional_ update or an update based on the old value of an attribute.
 The read and the write should be treated as a single transaction - protected
 (_isolated_) from other parallel writes to the same item.
 
-Alternator could do this isolation by using ScyllaDB's LWT (lightweight
+Alternator could do this isolation by using Scylla's LWT (lightweight
 transactions) for every write operation, but this significantly slows
 down writes, and not necessary for workloads which don't use read-modify-write
 (RMW) updates.
@@ -41,7 +41,7 @@ isolation policy for a specific table can be overridden by tagging the table
     which need a read before the write. An attempt to use such statements
     (e.g.,  UpdateItem with a ConditionExpression) will result in an error.
     In this mode, the remaining write requests which are allowed - pure writes
-    without a read - are performed using standard ScyllaDB writes, not LWT,
+    without a read - are performed using standard Scylla writes, not LWT,
     so they are significantly faster than they would have been in the
     `always_use_lwt`, but their isolation is still correct.
 
@@ -65,19 +65,19 @@ isolation policy for a specific table can be overridden by tagging the table
     read-modify-write updates. This mode is not recommended for any use case,
     and will likely be removed in the future.
 
-## Accessing system tables from ScyllaDB
-ScyllaDB exposes lots of useful information via its internal system tables,
+## Accessing system tables from Scylla
+Scylla exposes lots of useful information via its internal system tables,
 which can be found in system keyspaces: 'system', 'system\_auth', etc.
 In order to access to these tables via alternator interface,
 Scan and Query requests can use a special table name:
 `.scylla.alternator.KEYSPACE_NAME.TABLE_NAME`
-which will return results fetched from corresponding ScyllaDB table.
+which will return results fetched from corresponding Scylla table.
 
 This interface can be used only to fetch data from system tables.
 Attempts to read regular tables via the virtual interface will result
 in an error.
 
-Example: in order to query the contents of ScyllaDB's `system.large_rows`,
+Example: in order to query the contents of Scylla's `system.large_rows`,
 pass `TableName='.scylla.alternator.system.large_rows'` to a Query/Scan
 request.
 
@@ -113,14 +113,14 @@ connection (either active or idle), not necessarily an active request as
 in Alternator.
 
 ## Service discovery
-As explained in [ScyllaDB Alternator for DynamoDB users](compatibility.md),
+As explained in [Scylla Alternator for DynamoDB users](compatibility.md),
 Alternator requires a load-balancer or a client-side load-balancing library
-to distribute requests between all ScyllaDB nodes. This load-balancer needs
-to be able to _discover_ the ScyllaDB nodes. Alternator provides two special
+to distribute requests between all Scylla nodes. This load-balancer needs
+to be able to _discover_ the Scylla nodes. Alternator provides two special
 requests, `/` and `/localnodes`, to help with this service discovery, which
 we will now explain.
 
-Some setups know exactly which ScyllaDB nodes were brought up, so all that
+Some setups know exactly which Scylla nodes were brought up, so all that
 remains is to periodically verify that each node is still functional. The
 easiest way to do this is to make an HTTP (or HTTPS) GET request to the node,
 with URL `/`. This is a trivial GET request and does **not** need to be
@@ -133,10 +133,10 @@ $ curl http://localhost:8000/
 healthy: localhost:8000
 ```
 
-In other setups, the load balancer might not know which ScyllaDB nodes exist.
-For example, it may be possible to add or remove ScyllaDB nodes without a
+In other setups, the load balancer might not know which Scylla nodes exist.
+For example, it may be possible to add or remove Scylla nodes without a
 client-side load balancer knowing. For these setups we have the `/localnodes`
-request that can be used to discover which ScyllaDB nodes exist: A load balancer
+request that can be used to discover which Scylla nodes exist: A load balancer
 that already knows at least one live node can discover the rest by sending
 a `/localnodes` request to the known node. It's again an unauthenticated
 HTTP (or HTTPS) GET request:
@@ -160,7 +160,7 @@ list the nodes in a specific _data center_ or _rack_. These options are
 useful for certain use cases:
 
 * A `dc` option (e.g., `/localnodes?dc=dc1`) can be passed to list the
-  nodes in a specific ScyllaDB data center, not the data center of the node
+  nodes in a specific Scylla data center, not the data center of the node
   being contacted. This is useful when a client knowns of _some_ Scylla
   node belonging to an unknown DC, but wants to list the nodes in _its_
   DC, which it knows by name.
@@ -191,7 +191,7 @@ tells them to.
 
 If you want to influence whether a specific Alternator table is created with tablets or vnodes,
 you can do this by specifying the `system:initial_tablets` tag
-(in earlier versions of ScyllaDB the tag was `experimental:initial_tablets`)
+(in earlier versions of Scylla the tag was `experimental:initial_tablets`)
 in the CreateTable operation. The value of this tag can be:
 
 * Any valid integer as the value of this tag enables tablets.

docs/cql/cql-extensions.md

@@ -106,15 +106,6 @@ which is recommended in order to make the operation less heavyweight
 and allow for running multiple parallel pruning statements for non-overlapping
 token ranges.
 
-By default, the PRUNE MATERIALIZED VIEW statement is relatively slow, only
-performing one base read or write at a time. This can be changed with the
-USING CONCURRENCY clause. If the clause is used, the concurrency of reads
-and writes from the base table will be allowed to increase up to the specified
-value. For example, to run the PRUNE with 100 parallel reads/writes, you can use:
-```cql
-  PRUNE MATERIALIZED VIEW my_view WHERE v = 19 USING CONCURRENCY 100;
-```
-
 ## Synchronous materialized views
 
 Usually, when a table with materialized views is updated, the update to the

docs/cql/ddl.rst

@@ -365,7 +365,7 @@ Modifying a keyspace with tablets enabled is possible and doesn't require any sp
 
 - The replication factor (RF) can be increased or decreased by at most 1 at a time. To reach the desired RF value, modify the RF repeatedly.
 - The ``ALTER`` statement rejects the ``replication_factor`` tag. List the DCs explicitly when altering a keyspace. See :ref:`NetworkTopologyStrategy <replication-strategy>`.
-- An RF change cannot be requested while another RF change is pending for the same keyspace. Attempting to execute an ``ALTER`` statement in this scenario will fail with an explicit error. Wait for the ongoing RF change to complete before issuing another ``ALTER`` statement.
+- If there's any other ongoing global topology operation, executing the ``ALTER`` statement will fail (with an explicit and specific error) and needs to be repeated.
 - The ``ALTER`` statement may take longer than the regular query timeout, and even if it times out, it will continue to execute in the background.
 - The replication strategy cannot be modified, as keyspaces with tablets only support ``NetworkTopologyStrategy``.
 - The ``ALTER`` statement will fail if it would make the keyspace :term:`RF-rack-invalid <RF-rack-valid keyspace>`.
@@ -1043,8 +1043,6 @@ The following modes are available:
    * - ``immediate``
      - Tombstone GC is immediately performed. There is no wait time or repair requirement. This mode is useful for a table that uses the TWCS compaction strategy with no user deletes. After data is expired after TTL, ScyllaDB can perform compaction to drop the expired data immediately.
 
-.. warning:: The ``repair`` mode is not supported for :term:`Colocated Tables <Colocated Table>` in this version.
-
 .. _cql-per-table-tablet-options:
 
 Per-table tablet options

docs/cql/time-to-live.rst

@@ -102,7 +102,6 @@ Additional Information
 
 To learn more about TTL, and see a hands-on example, check out `this lesson <https://university.scylladb.com/courses/data-modeling/lessons/advanced-data-modeling/topic/expiring-data-with-ttl-time-to-live/>`_ on ScyllaDB University.
 
-* `Video: Managing data expiration with Time-To-Live <https://www.youtube.com/watch?v=SXkbu7mFHeA>`_
 * :doc:`Apache Cassandra Query Language (CQL) Reference </cql/index>`
 * :doc:`KB Article:How to Change gc_grace_seconds for a Table </kb/gc-grace-seconds/>`
 * :doc:`KB Article:Time to Live (TTL) and Compaction </kb/ttl-facts/>`

docs/dev/protocol-extensions.md

@@ -74,8 +74,6 @@ The keys and values are:
     as an indicator to which shard client wants to connect. The desired shard number
     is calculated as: `desired_shard_no = client_port % SCYLLA_NR_SHARDS`.
     Its value is a decimal representation of type `uint16_t`, by default `19142`.
-  - `CLIENT_OPTIONS` is a string containing a JSON object representation that
-    contains CQL Driver configuration, e.g. load balancing policy, retry policy, timeouts, etc.
 
 Currently, one `SCYLLA_SHARDING_ALGORITHM` is defined,
 `biased-token-round-robin`. To apply the algorithm,
@@ -238,26 +236,3 @@ the same mechanism for other protocol versions, such as CQLv4.
 
 The feature is identified by the `SCYLLA_USE_METADATA_ID` key, which is meant to be sent
 in the SUPPORTED message.
-
-## Sending the CLIENT_ROUTES_CHANGE event
-
-This extension allows a driver to update its connections when the
-`system.client_routes` table is modified.
-
-In some network topologies a specific mapping of addresses and ports is required (e.g.
-to support Private Link). This mapping can change dynamically even when no nodes are
-added or removed. The driver must adapt to those changes; otherwise connectivity can be
-lost.
-
-The extension is implemented as a new `EVENT` type: `CLIENT_ROUTES_CHANGE`. The event
-body consists of:
-- [string] change
-- [string list] connection_ids
-- [string list] host_ids
-
-There is only one change value: `UPDATE_NODES`, which means at least one client route
-was inserted, updated, or deleted.
-
-Events already have a subscription mechanism similar to protocol extensions (that is,
-the driver only receives the events it explicitly subscribed to), so no additional
-`cql_protocol_extension` key is introduced for this feature.

docs/dev/topology-over-raft.md

@@ -86,7 +86,6 @@ stateDiagram-v2
         de_left_token_ring --> [*]
     }
     state removing {
-        re_left_token_ring : left_token_ring
         re_tablet_draining : tablet_draining
         re_tablet_migration : tablet_migration
         re_write_both_read_old : write_both_read_old
@@ -99,8 +98,7 @@ stateDiagram-v2
         re_tablet_draining --> re_write_both_read_old
         re_write_both_read_old --> re_write_both_read_new: streaming completed
         re_write_both_read_old --> re_rollback_to_normal: rollback
-        re_write_both_read_new --> re_left_token_ring
-        re_left_token_ring --> [*]
+        re_write_both_read_new --> [*]
     }
     rebuilding --> normal: streaming completed
     decommissioning --> left: operation succeeded
@@ -124,10 +122,9 @@ Note that these are not all states, as there are other states specific to tablet
     Writes to vnodes-based tables are going to both new and old replicas (new replicas means calculated according
     to modified token ring), reads are using old replicas.
 - `write_both_read_new` - as above, but reads are using new replicas.
-- `left_token_ring` - the decommissioning or removing node left the token ring, but we still need to wait until other
-    nodes observe it and stop sending writes to this node. For decommission, we tell the node to shut down,
-    then remove it from group 0. For removenode, the node is already down, so we skip the shutdown step.
-    We also use this state to rollback a failed bootstrap or decommission.
+- `left_token_ring` - the decommissioning node left the token ring, but we still need to wait until other
+    nodes observe it and stop sending writes to this node. Then, we tell the node to shut down and remove
+    it from group 0. We also use this state to rollback a failed bootstrap or decommission.
 - `rollback_to_normal` - the decommission or removenode operation failed. Rollback the operation by
     moving the node we tried to decommission/remove back to the normal state.
 - `lock` - the topology stays in this state until externally changed (to null state), preventing topology
@@ -144,9 +141,7 @@ reads that started before this point exist in the system. Finally we remove the
 transitioning state.
 
 Decommission, removenode and replace work similarly, except they don't go through
-`commit_cdc_generation`. Both decommission and removenode go through the
-`left_token_ring` state to run a global barrier ensuring all nodes are aware
-of the topology change before the operation completes.
+`commit_cdc_generation`.
 
 The state machine may also go only through the `commit_cdc_generation` state
 after getting a request from the user to create a new CDC generation if the

docs/dev/view-building-coordinator.md

@@ -41,12 +41,12 @@ Unless the task was aborted, the worker will eventually reply that the task was
 it temporarily saves list of ids of finished tasks and removes those tasks from group0 state (pernamently marking them as finished) in 200ms intervals. (*)
 This batching of removing finished tasks is done in order to reduce number of generated group0 operations.
 
-On the other hand, view building tasks can can also be aborted due to 2 main reasons:
+On the other hand, view buildind tasks can can also be aborted due to 2 main reasons:
 - a keyspace/view was dropped
 - tablet operations (see [tablet operations section](#tablet-operations))
 In the first case we simply delete relevant view building tasks as they are no longer needed.
-But if a task needs to be aborted due to tablet operation, we're firstly setting the `aborted` flag to true. We need to do this because we need the task information
-to create new adjusted tasks (if the operation succeeded) or rollback them (if the operation failed).
+But if a task needs to be aborted due to tablet operation, we're firstly setting the `aborted` flag to true. We need to do this because we need the task informations
+to created a new adjusted tasks (if the operation succeeded) or rollback them (if the operation failed).
 Once a task is aborted by setting the flag, this cannot be revoked, so rolling back a task means creating its duplicate and removing the original task.
 
 (*) - Because there is a time gap between when the coordinator learns that a task is finished (from the RPC response) and when the task is marked as completed,