[SERVER-16441] syncSourceFeedback can spin, with network traffic, during reconfig Created: 05/Dec/14  Updated: 12/Jan/17  Resolved: 05/Dec/14

Status: Closed
Project: Core Server
Component/s: Replication
Affects Version/s: None
Fix Version/s: 2.8.0-rc3

Type: Bug Priority: Major - P3
Reporter: Eric Milkie Assignee: Eric Milkie
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Related
related to SERVER-26494 remove unreachable else-branch in syn... Closed
related to SERVER-27397 Disable OplogFetcher sync source re-e... Closed
is related to SERVER-16272 SyncSourceFeedback spams log on errors Closed
Operating System: ALL
Participants:

 Description   

During a replica set reconfig, each node independently increments and installs the new config. While this is happening, it is possible for a node to attempt to run an updatePosition command on a sync source node with a different config version than its own. Currently, the code tries again immediately until the config versions between the two nodes becomes the same and the updatePosition command finally succeeds; this can result in heavy network traffic.
One stopgap solution to this problem is to put in a sleep between retries.



 Comments   
Comment by Githook User [ 05/Dec/14 ]

Author:

{u'username': u'milkie', u'name': u'Eric Milkie', u'email': u'milkie@10gen.com'}

Message: SERVER-16441 treat updatePosition errors the same as handshake errors: add to blacklist.
Branch: master
https://github.com/mongodb/mongo/commit/3b3fe4973d16903f2817ce87f605b3bbe54d5a99

Generated at Thu Feb 08 03:41:03 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.