[SERVER-1516] if the primary fails, the remaining servers never sync up Created: 29/Jul/10  Updated: 12/Jul/16  Resolved: 30/Jul/10

Status: Closed
Project: Core Server
Component/s: Replication
Affects Version/s: None
Fix Version/s: 1.5.7

Type: Bug Priority: Major - P3
Reporter: Kristina Chodorow (Inactive) Assignee: Dwight Merriman
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Operating System: ALL
Participants:

 Description   

Test below. If you have 3 servers and overwhelm the master with inserts, then block it before the slaves have caught up (so, say one slave has 4000 docs and one slave has 3500 docs) they'll stay in that state forever. Seems like they should both end up with either 4000 or 3500 docs.

Run jstest/replsets/sync1.js to see this issue.



 Comments   
Comment by Dwight Merriman [ 30/Jul/10 ]

hopefully works now?

good find.

Comment by auto [ 30/Jul/10 ]

Author:

{'login': 'dwight', 'name': 'dwight', 'email': 'dwight@10gen.com'}

Message: SERVER-1516
http://github.com/mongodb/mongo/commit/7572badd921a2dfbcb7eb954cdb16c92ba9e92be

Comment by Dwight Merriman [ 30/Jul/10 ]

intermittent for me now

Comment by Kristina Chodorow (Inactive) [ 30/Jul/10 ]

Still doesn't work for me.

Comment by Dwight Merriman [ 30/Jul/10 ]

i think works now. there was a bad assertion. LMK

Comment by Kristina Chodorow (Inactive) [ 30/Jul/10 ]

sorry, network partition it. I deluged it with writes and then called

{replSetTest:1,blind:1}

before it had a chance to finish replicating.

Comment by Dwight Merriman [ 29/Jul/10 ]

what does "block it" mean?

Generated at Thu Feb 08 02:57:15 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.