-
Type:
Task
-
Resolution: Unresolved
-
Priority:
Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
None
-
Cluster Scalability
-
(copied to CRM)
Currently during the catch-up cloning phase of a chunk migration, batches of transferMods are fetched and replicated and each batch is currently hard coded to wait up to 180k iterations for the required writeConcern to be satisfied. This loop will insert numerous waiters in the replication waiterList. Instead of inserting so many entries in the waiterList, we should appropriately leverage future/promises to wait for replication during this stage of chunk migrations.
- is duplicated by
-
SERVER-89242 Unsatisfied write concern waiters can grow the replication waiter list unbounded
-
- Closed
-