-
Type: Bug
-
Resolution: Fixed
-
Priority: Major - P3
-
Affects Version/s: Backlog
-
Component/s: Sharding
-
Fully Compatible
-
ALL
-
140
Consider the following scenario:
- s0:n1 became primary
- s0:n1 starts a createCollectionCoordinator
- s0:n1 persist the coordinator document on disk
- s0:n1 aquire the recoverable critical section
- s0:n1 stepdowns just after committing the create collection
- s0:n2 resume the coordinator
- s0:n2 found the collection has been already created and exit without releasing the critical section
The goal of this ticket is to prevent the above scenario and re-enable the drop_collection_sharded.js FSM workload on stepdown suites
- is depended on by
-
SERVER-54905 Implement FSMs hierarchy testing concurrent DLL/CRUD/FCV operations
- Closed
-
SERVER-56555 Enable drop_collection_sharded.js FSM workload in stepdown suites
- Closed