WT-7685 fixed a unit issue related to max_latency. Since then, the threshold seems too high (1s) for insert/read/update operations, no warnings are raised at all.
The goal here is to reduce this threshold so at least a few warnings are raised. The investigation has been done in WT-7678 and the agreed baseline value for aws-perf is 60ms.
- depends on
-
WT-7678 Create a Jenkins perf job based on many-dhandle-stress workgen workload
- Closed