In jstestfuzz_replication_write_conflicts.yml, we enable the parameter that prints WriteConflictException stack traces for all operations here. There was a build failure where 10 concurrent clients were running, 5 of which were running collection.remove({}) simultaneously on a collection of 100,000 documents. This caused a lot of WriteConflictExceptions to be thrown between these operations. These operations had to print stack traces, which caused a slowdown leading to the 2-hour timeout.

When the WTWriteConflictException or WTWriteConflictExceptionForReads fail points are used and hit, we should consider printing the stack traces for those operations, as opposed to using traceWriteConflictExceptions. This would reduce the amount of noise and logging produced in the test.