[SERVER-16026] Documents with secondary unique indexed fields may not replicate properly on KV storage engines Created: 07/Nov/14  Updated: 11/Jul/16  Resolved: 10/Nov/14

Status: Closed
Project: Core Server
Component/s: Index Maintenance, Replication
Affects Version/s: 2.7.8
Fix Version/s: 2.8.0-rc0

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

Issue Links:
Duplicate
Operating System: ALL
Participants:
Linked BF Score: 0

 Description   

When replicating document ops with doc-level-locking storage engines, we may batch apply operations in parallel within the same collection. If unique secondary indexes exist (that is, unique indexes other than id), their unique constraints may be temporarily violated during the application of a batch of ops. If this happens, the secondary node would detect this and shut down before data corruption resulted.
In order to fix this, we need to disable unique key checking while applying ops on secondaries.



 Comments   
Comment by Githook User [ 10/Nov/14 ]

Author:

{u'username': u'andy10gen', u'name': u'Andy Schwerin', u'email': u'schwerin@mongodb.com'}

Message: SERVER-16026 Improved combination of ns and _id field into hash value for worker thread selection.
Branch: master
https://github.com/mongodb/mongo/commit/cb82efb115f2bca1cc5d7274d1f7e67b2326846e

Comment by Githook User [ 10/Nov/14 ]

Author:

{u'username': u'erh', u'name': u'Eliot Horowitz', u'email': u'eliot@10gen.com'}

Message: SERVER-16026: on secnodaries w/doc locking, secondary unique indexes always have to be unchecked during batch
Branch: master
https://github.com/mongodb/mongo/commit/c686f5a1949339ba9390e6efc01ef8421867c692

Generated at Thu Feb 08 03:39:43 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.