The fuzzer appears to have run killOp against the journal flusher thread. At least, right after there's a log about killing op with opId 8221, the journal flusher gets interrupted with the same error message as OperationContext::checkForInterruptNoAssert, which can be called from the locking code (the journal flusher takes a lock to do a write to the oplogTruncateAfterPoint), can return.
- related to
-
SERVER-79026 Failing to cancel the JournalFlusher thread might lead to 3-way deadlock
- Closed
-
SERVER-79810 make JournalFlusher::waitForJournalFlush() interruptible when waiting for write concern
- Closed