mirrors/scylladb
1
0
Fork 0
mirror of https://github.com/scylladb/scylladb.git synced 2026-05-12 19:02:12 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge 'make topology_coordinator::run noexcept' from Gleb

Browse Source 
Topology coordinator should handle failures internally as long as it
remains to be the coordinator. The raft state monitor is not in better
position to handle any errors thrown by it, all it can do it to restart
the coordinator. The series makes topology_coordinator::run handle all
the errors internally and mark the function as noexcept to not leak
error handling complexity into the raft state monitor.

* 'gleb/15728-fix' of github.com:scylladb/scylla-dev:
  storage_service: raft topology: mark topology_coordinator::run function as noexcept
  storage_service: raft topology: do not throw error from fence_previous_coordinator()
This commit is contained in:
Kamil Braun
2023-10-24 12:16:36 +02:00
parent 4abcec9296 dcaaa74cd4
commit 2a21029ff5
1 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
service/storage_service.cc
View File

@@ -2339,7 +2339,7 @@ class topology_coordinator {
co_await _topo_sm.event.when();
}
future<> fence_previous_coordinator();
future<> fence_previous_coordinator() noexcept;
public:
topology_coordinator(
@@ -2358,7 +2358,7 @@ public:
, _ring_delay(ring_delay)
{}
future<> run();
future<> run() noexcept;
};
future<bool> topology_coordinator::maybe_start_tablet_migration(group0_guard guard) {
@@ -2385,7 +2385,7 @@ future<bool> topology_coordinator::maybe_start_tablet_migration(group0_guard gua
co_return true;
}
future<> topology_coordinator::fence_previous_coordinator() {
future<> topology_coordinator::fence_previous_coordinator() noexcept {
// Write empty change to make sure that a guard taken by any previous coordinator cannot
// be used to do a successful write any more. Otherwise the following can theoretically happen
// while a coordinator tries to execute RPC R and move to state S.
@@ -2399,7 +2399,7 @@ future<> topology_coordinator::fence_previous_coordinator() {
// Topology state machine moves to state S while RPC R is still running.
// If RPC is idempotent that should not be a problem since second one executed by B will do nothing,
// but better to be safe and cut off previous write attempt
while (true) {
while (!_as.abort_requested()) {
try {
auto guard = co_await start_operation();
topology_mutation_builder builder(guard.write_timestamp());
@@ -2410,12 +2410,12 @@ future<> topology_coordinator::fence_previous_coordinator() {
continue;
} catch (...) {
slogger.error("raft topology: failed to fence previous coordinator {}", std::current_exception());
throw;
}
co_await seastar::sleep_abortable(std::chrono::seconds(1), _as);
}
}
future<> topology_coordinator::run() {
future<> topology_coordinator::run() noexcept {
slogger.info("raft topology: start topology coordinator fiber");
auto abort = _as.subscribe([this] () noexcept {