-
Type:
Task
-
Resolution: Done
-
Priority:
Major - P3
-
Affects Version/s: None
-
Component/s: Replication
-
None
-
Fully Compatible
-
v3.2
-
Repl 16 (06/24/16)
-
0
Currently, connections that are subjected to closing on stepdown are tagged in command::run(), this leaves the gaps before and after command::run() where closing could happen. Also, we have to consider retry if commands may hit the old primary during its stepping down.
Instead, this ticket suggests tagging all internal connections at the very beginning of establishing connections with an isMaster() call. We can add an extra field to isMaster to specify internal use so that servers may tag these connections KeepOpen.
- is related to
-
SERVER-24630 Mongos erroneously advances config optime for writes that fail write concern
-
- Closed
-
-
-
- Closed
-
-
SERVER-21456 Improve closing connection behavior when stepping down from primary
-
- Closed
-