Currently, it may happen that the last promoted index block includes
the partition_end marker. That's because we first write the partition
end marker and then emit the unclosed block. This behavior matches
Cassandra (checked in 3.x and 5.0.1).
This is problematic for ruling out data file reads based on index.
The width field is currently unused, but it will be used later where
the width of the last block is used to compute the skip position past
the last block for lookups which land after all keys in the
partition. If width includes the marker then such a skip would land in
the next partition, which is incorrect, as the reader context expects
a cell element. Even if that was recognized, it's wrong - if this is
not a single partition read (so upper bound is not at the next
partition too), then we would read from the wrong (next) partition.
We want to be able to make such skips in order to avoid unnecessary
data file IO for reads of missing rows. Currently, we would always
read the last block even if the key is past its "end" position.
Another way to solve this would be to propagate the "past the last
block" condition from the index cursor to the reader and let it deal
with it, but the logic for that would be complicated. With this fix,
there is no special logic required.
Single-row reads from large partition issue 64 KiB reads to the data file,
which is equal to the default span of the promoted index block in the data file.
If users would want to reduce selectivity of the index to speed up single-row reads,
this won't be effective. The reason is that the reader uses promoted index
to look up the start position in the data file of the read, but end position
will in practice extend to the next partition, and amount of I/O will be
determined by the underlying file input stream implementation and its
read-ahead heuristics. By default, that results in at least 2 IOs 32KB each.
There is already infrastructure to lookup end position based on upper
bound of the read, but it's not effective becasue it's a
non-populating lookup and the upper bound cursor has its own private
cached_promoted_index, which is cold when positions are computed. It's
non-populating on purpose, to avoid extra index file IO to read upper
bound. In case upper bound is far-enough from the lower bound, this
will only increase the cost of the read.
The solution employed here is to warm up the lower bound cursor's
cache before positions are computed, and use that cursor for
non-populating lookup of the upper bound.
We use the lower bound cursor and the slice's lower bound so that we
read the same blocks as later lower-bound slicing would, so that we
don't incur extra IO for cases where looking up upper bound is not
worth it, that is when upper bound is far from the lower bound. If
upper bound is near lower bound, then warming up using lower bound
will populate cached_promoted_index with blocks which will allow us to
locate the upper bound block accurately. This is especially important
for single-row reads, where the bounds are around the same key. In
this case we want to read the data file range which belongs to a
single promoted index block. It doesn't matter that the upper bound
is not exactly the same. They both will likely lie in the same block,
and if not, binary search will bring adjacent blocks into cache. Even
if upper bound is not near, the binary search will populate the cache
with blocks which can be used to narrow down the data file range
somewhat.
Fixes#10030.
The change was tested with perf-fast-forward.
I populated the data set with `column_index_size_in_kb` set to 1
scylla perf-fast-forward --populate --run-tests=large-partition-slicing --column-index-size-in-kb=1
Test run:
build/release/scylla perf-fast-forward --run-tests=large-partition-select-few-rows -c1 --keep-cache-across-test-cases --test-case-duration=0
This test reads two rows from the middle of a large partition (1M
rows), of subsequent keys. The first read will miss in the index file
page cache, the second read will hit.
Notice that before the change, the second read issued 2 aio requests worth of 64KiB in total.
After the change, the second read issued 1 aio worth of 2 KiB. That's because promoted index block is larger than 1 KiB.
I verified using logging that the data file range matches a single promoted index block.
Also, the first read which misses in cache is still faster after the change.
Before:
running: large-partition-select-few-rows on dataset large-part-ds1
Testing selecting few rows from a large partition:
stride rows time (s) iterations frags frag/s mad f/s max f/s min f/s avg aio aio (KiB) blocked dropped idx hit idx miss idx blk c hit c miss c blk allocs tasks insns/f cpu
500000 1 0.009802 1 1 102 0 102 102 21.0 21 196 2 1 0 1 1 0 0 0 568 269 4716050 53.4%
500001 1 0.000321 1 1 3113 0 3113 3113 2.0 2 64 1 0 1 0 0 0 0 0 116 26 555110 45.0%
After:
running: large-partition-select-few-rows on dataset large-part-ds1
Testing selecting few rows from a large partition:
stride rows time (s) iterations frags frag/s mad f/s max f/s min f/s avg aio aio (KiB) blocked dropped idx hit idx miss idx blk c hit c miss c blk allocs tasks insns/f cpu
500000 1 0.009609 1 1 104 0 104 104 20.0 20 137 2 1 0 1 1 0 0 0 561 268 4633407 43.1%
500001 1 0.000217 1 1 4602 0 4602 4602 1.0 1 2 1 0 1 0 0 0 0 0 110 26 313882 64.1%
(cherry picked from commit dfb339376aff1ed961b26c4759b1604f7df35e54)
Will be needed by the reader to jump to the current block even if we
already advanced to it before, when setting up the reader context.
We want to advance to lower bound earlier, before the praser skips to
the lower bound. We want that in order to set input stream data file
range based on index. If we didn't have access to the current block
and used the result from advance_to(), the parser will think we're
already in the block which has lower_bound when it attempts to skip,
and will not skip, falling back to scanning.
This fixes a use-after-free bug when parsing clustering key across
pages.
Also includes a fix for allocating section retry, which is potentially not safe (not in practice yet).
Details of the first problem:
Clustering key index lookup is based on the index file page cache. We
do a binary search within the index, which involves parsing index
blocks touched by the algorithm. Index file pages are 4 KB chunks
which are stored in LSA.
To parse the first key of the block, we reuse clustering_parser, which
is also used when parsing the data file. The parser is stateful and
accepts consecutive chunks as temporary_buffers. The parser is
supposed to keep its state across chunks.
In 93482439, the promoted index cursor was optimized to avoid
fully page copy when parsing index blocks. Instead, parser is
given a temporary_buffer which is a view on the page.
A bit earlier, in b1b5bda, the parser was changed to keep shared
fragments of the buffer passed to the parser in its internal state (across pages)
rather than copy the fragments into a new buffer. This is problematic
when buffers come from page cache because LSA buffers may be moved
around or evicted. So the temporary_buffer which is a view on the LSA
buffer is valid only around the duration of a single consume() call to
the parser.
If the blob which is parsed (e.g. variable-length clustering key
component) spans pages, the fragments stored in the parser may be
invalidated before the component is fully parsed. As a result, the
parsed clustering key may have incorrect component values. This never
causes parsing errors because the "length" field is always parsed from
the current buffer, which is valid, and component parsing will end at
the right place in the next (valid) buffer.
The problematic path for clustering_key parsing is the one which calls
primitive_consumer::read_bytes(), which is called for example for text
components. Fixed-size components are not parsed like this, they store
the intermediate state by copying data.
This may cause incorrect clustering keys to be parsed when doing
binary search in the index, diverting the search to an incorrect
block.
Details of the solution:
We adapt page_view to a temporary_buffer-like API. For this, a new concept
is introduced called ContiguousSharedBuffer. We also change parsers so that
they can be templated on the type of the buffer they work with (page_view vs
temporary_buffer). This way we don't introduce indirection to existing algorithms.
We use page_view instead of temporary_buffer in the promoted
index parser which works with page cache buffers. page_view can be safely
shared via share() and stored across allocating sections. It keeps hold to the
LSA buffer even across allocating sections by the means of cached_file::page_ptr.
Fixes#20766Closesscylladb/scylladb#20837
* github.com:scylladb/scylladb:
sstables: bsearch_clustered_cursor: Add trace-level logging
sstables: bsearch_clustered_cursor: Move definitions out of line
test, sstables: Verify parsing stability when allocating section is retried
test, sstables: Verify parsing stability when buffers cross page boundary
sstables: bsearch_clustered_cursor: Switch parsers to work with page_view
cached_file: Adapt page_view to ContiguousSharedBuffer
cached_file: Change meaning of page_view::_size to be relative to _offset rather than page start
sstables, utils: Allow parsers to work with different buffer types
sstables: promoted_index_block_parser: Make reset() always bring parser to initial state
sstables: bsearch_clustered_cursor: Switch read_block_offset() to use the read() method
sstables: bsearch_clustered_cursor: Fix parsing when allocating section is retried
Fixes#20862
With the change in 60af2f3cb2 the bookkeep
for buffer memory was changed subtly, the problem here that we would
shrink buffer size before we after flush use said buffer's size to
decrement the buffer_list_bytes value, previously inc:ed by the full,
allocated size. I.e. we would slowly grow this value instead of adjusting
properly to actual used bytes.
Test included.
Closesscylladb/scylladb#20886
During migration cleanup, there's a small window in which the storage
group was stopped but not yet removed from the list. So concurrent
operations traversing the list could work with stopped groups.
During a test which emitted schema changes during migrations,
a failure happened when updating the compaction strategy of a table,
but since the group was stopped, the compaction manager was unable
to find the state for that group.
In order to fix it, we'll skip stopped groups when traversing the
list since they're unused at this stage of migration and going away
soon.
Fixes#20699.
Signed-off-by: Raphael S. Carvalho <raphaelsc@scylladb.com>
Closesscylladb/scylladb#20798
Removes the update command from the setup command.
This is required because versions now are not strictly pinned in the poetry.lock file since Sphinx ScyllaDB Theme 1.8.
Closesscylladb/scylladb#20876
To reduce the amount of space needed for reports, this PR will modify logs
attachment in allure, so it will attach logs only for the tests that have
status other than PASSED. To simplify the solution, with the current way it's
not possible to switch off these logs completely.
Closesscylladb/scylladb#20786
following headers are no longer used by this compilation unit:
- "utils/managed_ref.hh"
- "test/perf/perf.hh"
this was identified by clang-include-cleaner. As the code is audited,
we can safely remove the #include directive.
Signed-off-by: Kefu Chai <kefu.chai@scylladb.com>
Closesscylladb/scylladb#20850
since Python 3.13, passing count to `re.sub()` as positional argument
has been deprecated. and when runnint `test.py` with Python 3.13, we
have following warning:
```
/home/kefu/dev/scylladb/./test.py:1477: DeprecationWarning: 'count' is passed as positional argument
args.tests = set(re.sub(r'.* List configured unit tests\n(.*)\n', r'\1', out, 1, re.DOTALL).split("\n"))
```
see also https://github.com/python/cpython/issues/56166
in order to silence this distracting warning, let's pass
`count` using kwarg.
Signed-off-by: Kefu Chai <kefu.chai@scylladb.com>
Closesscylladb/scylladb#20859
Currently, node ops virtual task gathers its children from all nodes contained
in a sum of service::topology::normal_nodes and service::topology::transition_nodes.
The maps may contain nodes that are down but weren't removed yet. So, if a user
requests the status of a node ops virtual task, the task's attempt to retrieve
its children list may fail with seastar::rpc::closed_error.
Filter out the tasks that are down in node_ops::task_manager_module::get_nodes.
Fixes: #20843.
Closesscylladb/scylladb#20856
clang-tidy warns:
```
Warning: /__w/scylladb/scylladb/utils/directories.cc:132:52: warning: 'path' used after it was moved [bugprone-use-after-move]
132 | bool can_access = co_await file_accessible(path.string(), access_flags::read | access_flags::write | access_flags::execute);
| ^
/__w/scylladb/scylladb/utils/directories.cc:121:28: note: move occurred here
121 | verification_error(std::move(path), "File not owned by current euid: {}. Owner is: {}", geteuid(), sd.uid);
| ^
```
because we pass `std::move(path)` to `verification_error()`, and "then" use this variable again in this same function.
this is a false alarm, but we could make it very clear to convince this tool that it's safe to do so.
---
it's a cleanup, hence no need to backport.
Closesscylladb/scylladb#20875
* github.com:scylladb/scylladb:
directories: mark verification_error() with [[noreturn]]
directories: pass const ref of path to verification_error()
Reduce compile time and unnecessary compilations by reducing #include load.
Minor refactoring, no backport.
Closesscylladb/scylladb#20864
* github.com:scylladb/scylladb:
raft_group0_client: uninclude "raft_group0_registry.hh"
raft_group_registry: extract raft_timeout
raft_group0_client: uninclude "mutation/mutation.hh"
raft_group0_client: uninclude "db/system_keyspace.hh"
db: system_keyspace: extract auth_version_t into its own header
this helps the compiler or static analyzers do make the right decision.
for instance, clang-tidy thinks a parameter like `std::move(path)`
could be reused after being moved away. with this attribute, this tool
should be able to tell that this never happens.
Signed-off-by: Kefu Chai <kefu.chai@scylladb.com>
before this change, we pass a `path` to `verification_error()` by
moving away from the original `path`. this works fine in the sense
that it is correct and does not incur potential performance issues.
but clang-tidy considers it a used-after-move, because it cannot tell
`verification_error()` does not return at all, and believes that `path`
could be accessed again after being moved away. so it warns like:
```
Warning: /__w/scylladb/scylladb/utils/directories.cc:132:52: warning: 'path' used after it was moved [bugprone-use-after-move]
132 | bool can_access = co_await file_accessible(path.string(), access_flags::read | access_flags::write | access_flags::execute);
| ^
/__w/scylladb/scylladb/utils/directories.cc:121:28: note: move occurred here
121 | verification_error(std::move(path), "File not owned by current euid: {}. Owner is: {}", geteuid(), sd.uid);
| ^
```
in this change, instead of passing `fs::path` to `verification_error()`,
we pass a `const fs::path&` to this function. because
`verification_error()` is not coroutine, neither does it not pass `path` to
another continuation to be scheduled. so it's perfectly fine to pass
`path` to it.
this change address the false alarms from clang-tidy.
Signed-off-by: Kefu Chai <kefu.chai@scylladb.com>
While documenting materialized view in a new document (Refs #16569)
I encountered a few questions and this patch contains tests that
clarify their answer - and can later guarantee that the answer doesn't
unintentionally change in the future. The questions that these tests
answer are:
1. It is not allowed to filter a view on a static column (a comment
on the test explains why).
2. We already tested that it's not allowed to SELECT a static column into
a view. Here we add the check that "SELECT *" is also not allowed if
a static column exists in the base table.
3. We check that CREATE MATERIALIZED VIEW ... WITH COMMENT='..' works.
4. We check that CREATE MATERIALIZED VIEW ... WITH COMPACT STORAGE is
forbidden.
5. We check that CREATE MATERIALIZED VIEW ... WITH garbage=.. fails
with a clean InvalidRequest.
All these tests pass on both Scylla and Cassandra.
Signed-off-by: Nadav Har'El <nyh@scylladb.com>
Closesscylladb/scylladb#20873
One of the design goals of the test/cql-pytest frameworks was to be able
to run these tests against Cassandra. Preferably, we should be able to
run most of the tests against any popular version of Cassandra, including
Cassandra 3. This is admittingly a very old version, but was still maintained
until just a year ago, it's the version that Scylla is most compatible with,
and we can still be curious about how it worked.
Until recently cql-pytest indeed worked on Cassandra 3, but it broke on some
change related to tablet detection that cause our most basic fixture -
"text_keyspace" - to use the Cassandra 4 feature of "auto expand".
This is trivial to fix - we should just use the this_dc fixture that we already
had exactly for this purpose.
Fixes#20781
Signed-off-by: Nadav Har'El <nyh@scylladb.com>
Closesscylladb/scylladb#20782
* ./seastar 69f88e2f...3c9c2696 (14):
> core/reactor: don't check AIO block count when they are not needed
> build: do not print the default value of --c++-standard in help output
> json_formatter: Add tests for formatter::write
> Add APIs to get group details and to change ownership of file.
> scripts/perftune.py: improve a dry-run printout
> build: drop the workaround for a GCC bug
> cmake: Depend on libbsd if DPDK depends on it
> http: clarify the ownership in the router's doxygen comment
> build: check for P2582R1 support
> python: introduce a python formatting CI check
> addr2line: reformat with black
> scripts: add pyproject.toml
> json_formatter: Make formatter::write work for std::pair
> README.md: use the github homepage of Ceph for Crimson
Closesscylladb/scylladb#20836
On RHEL9, systemd-coredump fails to coredump on /var/lib/scylla/coredump because the service only have write acess with systemd_coredump_var_lib_t. To make it writable, we need to add file context rule for /var/lib/scylla/coredump, and run restorecon on /var/lib/scylla.
Fixes#19325Closesscylladb/scylladb#20528
* github.com:scylladb/scylladb:
scylla_raid_setup: configure SELinux file context
scylla_coredump_setup: fix SELinux configuration for RHEL9
It doesn't need it apart from a forward declaration.
Files that lost necessary includes are adjusted, and some users
of auth_version_t are redirected to the definition outside system_keyspace.
When `property_file` is provided, we generate a
`cassandra-rackdc.properties` file, but to actually use it,
`endpoint_snitch` must be set to `GossipingPropertyFileSnitch`.
Signed-off-by: Benny Halevy <bhalevy@scylladb.com>
Closesscylladb/scylladb#20730
In order to later use the formatter for the inner class
promoted_index_block, which is defined out of line after
cached_promoted_index class definition.
This fixes a use-after-free bug when parsing clustering key across
pages.
Clustering key index lookup is based on the index file page cache. We
do a binary search within the index, which involves parsing index
blocks touched by the algorithm. Index file pages are 4 KB chunks
which are stored in LSA.
To parse the first key of the block, we reuse clustering_parser, which
is also used when parsing the data file. The parser is stateful and
accepts consecutive chunks as temporary_buffers. The parser is
supposed to keep its state across chunks.
In b1b5bda, the parser was changed to keep shared fragments of the
buffer passed to the parser in its internal state (across pages)
rather than copy the fragments into a new buffer. This is problematic
when buffers come from page cache because LSA buffers may be moved
around or evicted. So the temporary_buffer which is a view on the LSA
buffer is valid only around the duration of a single consume() call to
the parser.
If the blob which is parsed (e.g. variable-length clustering key
component) spans pages, the fragments stored in the parser may be
invalidated before the component is fully parsed. As a result, the
parsed clustering key may have incorrect component values. This never
causes parsing errors because the "length" field is always parsed from
the current buffer, which is valid, and component parsing will end at
the right place in the next (valid) buffer.
The problematic path for clustering_key parsing is the one which calls
primitive_consumer::read_bytes(), which is called for example for text
components. Fixed-size components are not parsed like this, they store
the intermediate state by copying data.
This may cause incorrect clustering keys to be parsed when doing
binary search in the index, diverting the search to an incorrect
block.
The solution is to use page_view instead of temporary_buffer, which
can be safely shared via share() and stored across allocating
section. The page_view maintains its hold to the LSA buffer even
across allocating sections.
Fixes#20766
Currently, parsers work with temporary_buffer<char>. This is unsafe
when invoked by bsearch_clustered_cursor, which reuses some of the
parsers, and passes temporary_buffer<char> which is a view onto LSA
buffer which comes from the index file page cache. This view is stable
only around consume(). If parsing requires more than one page, it will
continue with a different input buffer. The old buffer will be
invalid, and it's unsafe for the parser to store and access
it. Unfortunetly, the temporary_buffer API allows sharing the buffer
via the share() method, which shares the underlying memory area. This
is not correct when the underlying is managed by LSA, because storage
may move. Parser uses this sharing when parsing blobs, e.g. clustering
key components. When parsing resumes in the next page, parser will try
to access the stored shared buffers pointing to the previous page,
which may result in use-after-free on the memory area.
In prearation for fixing the problem, parametrize parsers to work with
different kinds of buffers. This will allow us to instantiate them
with a buffer kind which supports sharing of LSA buffers properly in a
safe way.
It's not purely mechanical work. Some parts of the parsing state
machine still works with temporary_buffer<char>, and allocate buffers
internally, when reading into linearized destination buffer. They used
to store this destination in _read_bytes vector, same field which is
used to store the shared buffers. Now it's not possible, since shared
buffer type may be different than temporary_buffer<char>. So those
paths were changed to use a new field: _read_bytes_buf.
When reset() is done due to allocating section retry, it can be
theoretically in an arbitrary point. So we should not assume that it
finished parsing and state was reset by previous parsing. We should
reset all the fields.
Collection stress tests include testing of B- B+- and radix trees, and those tests live in unit/ suite. There are also small corner-case tests for those collections in boost/ suite. There's an attempt to get rid of unit suite in favor of boost one, and this PR moves the collections stress testing from unit suite into their boost counterparts.
refs: scylladb/qa-tasks#1655Closesscylladb/scylladb#20475
* github.com:scylladb/scylladb:
test: Move other collection-testing headers from unit to boost
test: Move stress-collecton header from unit to boost
test: Move B+tree compactiont test from unit to boost
test: Move radix tree compactiont test from unit to boost
test: Move B-tree compactiont test from unit to boost
test: Move radix tree stress test from unit to boost
test: Move B-tree stress test from unit to boost
test: Move b+tree stress test from unit to boost
test: Add bool in_thread argument to stress_collection function
The datadir keeps path to directory where local sstables can be. The very same information is now kept in table's storage options (#20542). This set fixes the remaining places that still use table::config::datadir and table::dir() and removes the datadir field.
Closesscylladb/scylladb#20675
* github.com:scylladb/scylladb:
treewide: Remove table::config::datadir
distributed_loader: Print storage options, not datadir
data_dictionary: Add formatter for storage_options
test: Construct table_for_tests with table storage options
test: Generalize pair of make_table_for_tests helpers
tests: Add helper to get snapshot directory from storage options
table: snapshot_exists: Get directory from storage options
table: snapshot_on_all_shards: Get directory from storage options
For each new node added to the raft config populate its ID to IP mapping
in raft address map from the gossiper. The mapping may have expired if a
node is added to the raft configuration long after it first appears in
the gossiper.
Fixesscylladb/scylladb#20600
Backport to all supported versions since the bug may cause bootstrapping failure.
Closesscylladb/scylladb#20601
* github.com:scylladb/scylladb:
test: extend existing test to check that a joining node can map addresses of all pre-existing nodes during join
group0: make sure that address map has an entry for each new node in the raft configuration
Parser's state was not reset when allocating section was retried.
This doesn't cause problems in practice, because reserves are enough
to cover allocation demands of parsing clustering keys, which are at
most 64K in size. But it's still potentially unsafe and needs fixing.
After commit d16ea0af, compiling the server using cmake fails with the
following error :
```
FAILED: service/CMakeFiles/service.dir/Dev/qos/service_level_controller.cc.o
...
/home/Scylla/scylladb/cql3/util.hh:21:10: fatal error: 'cql3/CqlParser.hpp' file not found
21 | #include "cql3/CqlParser.hpp"
| ^~~~~~~~~~~~~~~~~~~~
1 error generated.
```
Fix it by linking the cql3 to the service library.
Closesscylladb/scylladb#20805
Maintainers use scripts/pull_github_pr.sh from scylladb.git when merging PRs and before pushing to the next. We want to prevent merges from piling up on top of unstable builds. This change will check Gating's current status and notify the maintainers
Related to scylladb/scylla-pkg#3644Closesscylladb/scylladb#20742
When users create a table using the Alternator API, they can decide if the billing is PROVISIONED of PAY_PER_REQUEST.
If the billing is set to PROVISIONED, they need to set the ProvisionedThroughput ReadCapacityUnits (RCU) and WriteCapacityUnits (WCU).
This series adds support for getting and setting the ProvisionedThroughput. The values will be stored as table extension tags.
Following how TTL is stored within the Alternator, we will use ```system:rcu_attribute``` and ```system:wcu_attribute``` for the labels.
The series adds a test that sets ProvisionedThroughput and validates that it gets the value back. It was tested with both Alternator and AWS.
This series is part of the effort to monitor, limit, and bill Alternator operations.
New code, no need to backport.
Closesscylladb/scylladb#20056
* github.com:scylladb/scylladb:
docs/alternator/compatibility.md: explain the consumed capacity provisioned
Add test/alternator/test_provisioned_throughput.py
test/alternator/util.py: Allow override BillingMode
alternator/executor.cc: Store ProvisionedThroughput
A global index has a primary key of the form
(indexed_column, token, partition_key_column..., clustering_key_column...)
The primary key columns are used to point at the base table row, and
the token (computed as token(partition_key_column...) is used to maintain
sort order.
The query planner has an optimization: if the partition key is fully
constrained to a unique value, then we compute the token from the partition
key and use that to seek directly into the clustering row range for
that base table partition. If the clustering key is also partially
constrained, it is used to refine the index clustering key.
Currently, this optimization is implemented as a hack: the partition key
is extracted from the prepared statement + query options in
get_global_index_token_clustering_ranges(), then used to calculate
the token, which is then substituted in the expression passed to
get_single_column_clustering_bounds() (the expression is shared across
all running queries, so this is quite dangerous).
We simplify the whole thing:
- Let prepare_index_global() recognize that if the partition key is not
fully constrained, then there is no way that we'll be able to compute
the token (as it needs all partition key columns). Since the token
is the first clustering key column of the index table, we can truncate
it to length zero and bail out.
- Otherwise, the partition key is fully constrained. We refactor the
predicate (pk1 = :a AND pk2 = :b) to (pk1, pk2) := (:a, :b). We then
pass expressions representing the partition key to the token function,
ending up with token(:a, :b). We then substitute this expression into
(*_idx_tbl_ck_prefix)[0], which computes the first clustering key
column for the index table.
- Remove the runtime component in get_global_index_clustering_ranges().
Note this include the early return if the partition key wasn't fully
constrained (though the comment only mentions over-constraining), and
the token computation, which is now done by evaluate().
Closesscylladb/scylladb#20733
