[SERVER-17852] Reduce unnecessary replicated journal write concern flag for multi-update/delete Created: 01/Apr/15  Updated: 14/Apr/16  Resolved: 29/Apr/15

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

Type: Improvement Priority: Major - P3
Reporter: Siyuan Zhou Assignee: Siyuan Zhou
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Related
is related to SERVER-5218 Add a write concern to wait for journ... Closed
Backwards Compatibility: Fully Compatible
Participants:

 Description   

In implementing SERVER-5218, j:true will be respected on replicas, for multi-update and multi-delete, a j:true is added for all document updates generated. This might introduce unnecessary journaling waiting on all batches of ops of multi-update/delete on replicas, since waiting for only the batch including the last op is sufficient.



 Comments   
Comment by Scott Hernandez (Inactive) [ 01/Apr/15 ]

One possible implementation would be to introduce a new write concern operation type for replication to enforce the journal write at the end of the multi-operation or write command batch, instead of using the current "j" flag on an oplog entry.

This would allow us to be more precise with placing the "journal" event in the sequence of writes which are replicated, via the oplog.

Generated at Thu Feb 08 03:45:48 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.