In WouldChangeOwningShard tests it is possible to have commands fail due to Lock timeout due to a shard version mismatch refresh happening right before the write which prevents us from taking the collection lock in time. See SERVER-61461 and SERVER-67536 for other instances of this failure. We should increase the lock timeout in this test to prevent this from happening
- is related to
-
SERVER-69234 Make txn_recover_decision_using_recovery_router.js robust to spurious shard version refreshes
- Closed