thebestvibedev/tromcho.net
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Always continue operations to not miss resuming after the lack of PG primary

Browse Source 
Should fix spurious client hangs during PG primary switchover
This commit is contained in:
Vitaliy Filippov
2024-07-22 00:27:35 +03:00
parent da73d5f45a
commit 8635e0af24
1 changed files with 3 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/client/cluster_client.cpp
+3 -4
View File
@@ -452,10 +452,9 @@ void cluster_client_t::on_change_pg_state_hook(pool_id_t pool_id, pg_num_t pg_nu
if (pg_cfg.cur_primary != prev_primary) if (pg_cfg.cur_primary != prev_primary)
{ {
// Repeat this PG operations because an OSD which stopped being primary may not fsync operations // Repeat this PG operations because an OSD which stopped being primary may not fsync operations
if (wb->repeat_ops_for(this, 0, pool_id, pg_num) > 0) wb->repeat_ops_for(this, 0, pool_id, pg_num);
{ // Always continue to resume operations hung because of lack of the primary OSD
continue_ops(); continue_ops();
}
} }
} }