[SERVER-20604] CSRS with protocol version 1 is slow when doing reads from secondaries Created: 24/Sep/15  Updated: 19/May/17  Resolved: 06/Oct/15

Status: Closed
Project: Core Server
Component/s: Replication, Sharding
Affects Version/s: None
Fix Version/s: 3.1.9

Type: Bug Priority: Critical - P2
Reporter: Spencer Brody (Inactive) Assignee: Eric Milkie
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Depends
is depended on by SERVER-20722 Require journaling and rs pv1 with co... Closed
Related
related to SERVER-20800 getMore command against awaitData cur... Closed
related to SERVER-20493 CSRS with protocolVersion:0 is very s... Closed
related to SERVER-29275 Two Phase Drops: implement collection... Closed
is related to SERVER-20749 Race during getmore batch creation an... Closed
Backwards Compatibility: Fully Compatible
Operating System: ALL
Sprint: RPL A (10/09/15)
Participants:

 Comments   
Comment by Githook User [ 08/Oct/15 ]

Author:

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

Message: SERVER-20604 fix getMoreRequest.toBSON
Branch: master
https://github.com/mongodb/mongo/commit/fcb0320edf3a82ab644433319eb717db29532485

Comment by Githook User [ 06/Oct/15 ]

Author:

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

Message: SERVER-20604 unblock AwaitData queries if commit level needs updating downstream.
Branch: master
https://github.com/mongodb/mongo/commit/a8e6c0ebc70fad95a5e05281076cfa4b6db500c2

Comment by Eric Milkie [ 02/Oct/15 ]

Also, we need to send up the current known commit level, so that we can return immediately from getMore and not block, if the commit level has already advanced. Otherwise, we will block waiting for the AwaitData timeout and won't update the commit level in a timely manner.

Comment by Eric Milkie [ 02/Oct/15 ]

This is partly due to AwaitData delaying the propagation of the commit level down to the secondaries from the primary. If we can signal the AwaitData to return immediately when the commit level changes, this should go a long way to speeding up reads from secondaries.

Comment by Spencer Brody (Inactive) [ 24/Sep/15 ]

We should revert this line once this has been fixed

Generated at Thu Feb 08 03:54:43 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.