storage: GC queue tuning #5252
Labels
C-performance
Perf of queries or internals. Solution not expected to change functional behavior.
S-1-stability
Severe stability issues that can be fixed by upgrading, but usually don’t resolve by restarting
Milestone
The GC queue currently runs at regular intervals but doesn't actually get to delete much data. We should tune the threshold at which a replica is queued for GC (which is currently at one normalized units; too low).
To observe this, run the photos app for a while (single node is fine), then let the cluster idle and observe regular CPU spikes and the gc queue's event log.
The text was updated successfully, but these errors were encountered: