On a secondary node, the currentOp command includes storage stats in its slow op log. If oplog application is taking too long to apply a batch of operations, this may prevent the currentOp from completing in a timely manner.
=========
The defect addressed here primarily involves the Parallel Batch Writer Mutex (PBWM), a logical resource that is required for oplog application on a secondary node. Prior to this fix, it was possible for the database profiler and the currentOp command to compete for the PBWM resource with secondary oplog application.
- is related to
-
SERVER-38341 Remove Parallel Batch Writer Mutex
-
- Closed
-
- related to
-
SERVER-44822 update indexing tests to avoid logging slow operations
-
- Closed
-
-
-
- Closed
-
-
-
- Closed
-
-
-
- Closed
-
-
SERVER-76851 remove unnecessary latch predicate in index_commit_currentop_slow.js
-
- Closed
-