-
Type: Improvement
-
Resolution: Duplicate
-
Priority: Major - P3
-
None
-
Affects Version/s: None
-
Component/s: Sharding
-
None
Currently, the network call retrying logic for NotMaster and network error is scattered across multiple components - replica set monitor, shard registry, catalog manager and the async results merger.
We should come up with a comprehensive solution, which unifies the retries that we do in all these places.
- duplicates
-
SERVER-23647 Create helper function for encapsulating retry logic when talking to shards
- Closed
- related to
-
SERVER-21271 Cleanup the runCommand methods on ShardRegistry
- Closed