Details
-
New Feature
-
Status: Open (View Workflow)
-
Major
-
Resolution: Unresolved
-
None
-
None
Description
It is currently very hard to diagnose slave lag problems. This ticket is to track work to extend the replica with various diagnostic capabilities to make it easier to track down performance issues.
One idea so far is to have dynamic profiling that only turns on when the slave starts lagging. For example, it could use the perf C API; or automatically capture thread stacks.