-
Type:
Bug
-
Resolution: Fixed
-
Priority:
Major - P3
-
Affects Version/s: 5.0.0, 6.0.0, 7.0.0, 8.0.0
-
Component/s: None
-
Cluster Scalability
-
Fully Compatible
-
ALL
-
v8.2, v8.0, v7.0
-
None
-
None
-
None
-
None
-
None
-
None
-
None
If the range deleter encounters a task to delete a range of a collection with a hashed shard key but without a hashed shard key index, it will eventually skip it and reinsert it at the end of the queue. If that is the only task remaining in the queue, the main loop of the range deleter effectively degenerates to a tight loop where it is skipping and reinserting tasks to the queue as fast as the CPU allows, hogging the CPU unnecessarily.
- is caused by
-
SERVER-79632 Stop range deletion when hashed shard key index does not exist
-
- Closed
-
- is duplicated by
-
SERVER-110424 Rangedeletion just spins aggressively if only task it has cannot be completed due to missing index
-
- Closed
-
- is related to
-
SERVER-110456 Add check for range deleter tasks before allowing dropIndex on hashed shard key index
-
- Backlog
-
- related to
-
SERVER-71436 Range deleter must not aggressively spam the log when shard key index not found
-
- Closed
-