-
Type:
Bug
-
Resolution: Fixed
-
Priority:
Major - P3
-
Affects Version/s: None
-
Component/s: Sharding
-
Fully Compatible
-
ALL
-
Sharding 2021-04-05
-
163
-
1
It is possible for recipients to be in RecipientStateEnum::kError while documents are being copied, oplog entries are being fetched, and oplog entries are being applied. This is because the RecipientStateMachine will transition to kError as soon as any data replication component encounters an operation-fatal error. The data replication components will be shut down after the RecipientStateMachine transitions to kError.
- https://github.com/mongodb/mongo/blob/6afd283045dc43c0b249424ba03a72b89e5ad8c6/src/mongo/db/s/resharding/resharding_metrics.cpp#L174
- https://github.com/mongodb/mongo/blob/6afd283045dc43c0b249424ba03a72b89e5ad8c6/src/mongo/db/s/resharding/resharding_metrics.cpp#L185-L187
- https://github.com/mongodb/mongo/blob/6afd283045dc43c0b249424ba03a72b89e5ad8c6/src/mongo/db/s/resharding/resharding_metrics.cpp#L195-L196
should all be updated to include RecipientStateEnum::kError as well.