[SERVER-5183] Get rid of setNoJournal Created: 02/Mar/12  Updated: 11/Jul/16  Resolved: 22/Mar/12

Status: Closed
Project: Core Server
Component/s: Replication
Affects Version/s: None
Fix Version/s: 2.1.1

Type: Improvement Priority: Major - P3
Reporter: Mathias Stearn Assignee: Andy Schwerin
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-5101 Test failure in closeall.js Closed
Participants:

 Description   

Actually pretty easy and I think it will fix part of the closeall.js test. Code review should be up soon.



 Comments   
Comment by auto [ 22/Mar/12 ]

Author:

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

Message: fix local.slaves updating after SERVER-5183 fallout
not an ideal solution but should make tests pass
Branch: master
https://github.com/mongodb/mongo/commit/0950cc1f03b12cf1e27222825ab196ae7a9ecc0f

Comment by Eliot Horowitz (Inactive) [ 21/Mar/12 ]

I think Any change will get it reset.
But basically yes

Comment by Eliot Horowitz (Inactive) [ 21/Mar/12 ]

Yes. local.slaves is very important and we need the cache reset.
Therefore we have to filter it out when done from the updates thread.
I Can do later if you're not sure.

Comment by Andy Schwerin [ 21/Mar/12 ]

Will work on it. Are you certain that anyone actually consults local.slaves?

Comment by Mathias Stearn [ 21/Mar/12 ]

We need to either not clear cache on changes to local.slaves or filter out the changes due to cache flushing. I know how to do this, but I'll be on various planes for the next 24 hours or so (actually on a plane right now delayed at the gate.) Feel free to fix in the mean time.

Comment by Eric Milkie [ 21/Mar/12 ]

This seems to have possibly broke getlasterror/write concern.
In particular, replsets/initial_sync3.js and replsets/majority.js fail mostly consistently, because the slave cache is getting cleared before the getlasterror logic can successfully see that all slaves have synced the oplog enough.
By setting (hacking) the delay in the slavetracker loop up really high, it makes the test pass again, because the cache isn't getting cleared as frequently...

Comment by auto [ 14/Mar/12 ]

Author:

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

Message: no one uses setNoJournal anymore so get rid of SERVER-5183
Branch: master
https://github.com/mongodb/mongo/commit/ced3eb63cf349aa42bcffaa8a8d06ee445e0ca60

Comment by auto [ 14/Mar/12 ]

Author:

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

Message: Make SlaveTracking::update not update row directly SERVER-5183
Branch: master
https://github.com/mongodb/mongo/commit/6d9beb1b39b24a98014ddb586cddfa893ccf2a44

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