[SERVER-8378] indexbg1.js failures Created: 29/Jan/13  Updated: 11/Jul/16  Resolved: 29/Oct/13

Status: Closed
Project: Core Server
Component/s: Index Maintenance, Testing Infrastructure
Affects Version/s: 2.3.2
Fix Version/s: 2.5.4

Type: Bug Priority: Major - P3
Reporter: Eric Milkie Assignee: Matt Dannenberg
Resolution: Done Votes: 0
Labels: build-failure, buildbot
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Duplicate
is duplicated by SERVER-10505 slowWeekly test indexbg1.js has been ... Closed
Related
related to SERVER-8579 Consolidate Mongod Lock/Resource Sche... Closed
Operating System: ALL
Participants:

 Description   

slowWeekly/indexbg1.js is stalling on the first write (which is a remove)
while a background index is going.
There is also a race condition on fast machines where it doesn't really wait for the background index to start, which is why the test passes on Linux 64.



 Comments   
Comment by auto [ 29/Oct/13 ]

Author:

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

Message: SERVER-8378: make indexbg1 more reliable by not making it do multiple table scans
Branch: master
https://github.com/mongodb/mongo/commit/202fe3e238fc378656482291d724728786581a2b

Comment by auto [ 18/Oct/13 ]

Author:

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

Message: SERVER-8378 reactivate parts of test, since new query engine
Branch: master
https://github.com/mongodb/mongo/commit/363c19f1a8b14398128d81ba7a6db7de058298b0

Comment by Eric Milkie [ 15/Oct/13 ]

Yes we should as soon as we cut 2.5.3.

Comment by Matt Kangas [ 14/Oct/13 ]

milkie still need to turn this test back on?

Comment by Eric Milkie [ 30/Aug/13 ]

Need to keep this open until we revert some of the changes I made to this test, so that we can test doing deletes while a bg index is running.

Comment by auto [ 30/Aug/13 ]

Author:

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

Message: SERVER-8378 make bg index test more reliable

I needed to comment out the 'remove' part because the current deleteObject code gets bullied if
another writer is taking a write lock for long quanta, and it never makes forward progress.
Hopefully the new query planner code, which uses a better yielding strategy, will allow
this to be turned back on soon.
Branch: master
https://github.com/mongodb/mongo/commit/bd8ecf1eddc777d5763532920a20f442690a8545

Comment by Matt Kangas [ 30/Aug/13 ]

SERVER-10505 "slowWeekly test indexbg1.js has been failing on OS X builders since January"

Still seen today, August 30.

osx-108-dur-off slow_weekly_tests : indexbg1.js
osx-108 slow_weekly_tests : indexbg1.js

"assert failed : unable to run checks in parallel with index creation"
http://mci.10gen.com/ui/task/mongodb_mongo_master_osx_108_dur_off_84ff1694c4bf92f3deef5c013ac9a5b912e6f96f_13_08_29_22_09_07_slow_weekly_tests_osx_108
http://buildlogs.mongodb.org/mci_0.9_osx_108_dur_off/builds/675/test/slow_weekly_tests_0/indexbg1.js

"assert failed : unable to run checks in parallel with index creation"
http://mci.10gen.com/ui/task/mongodb_mongo_master_osx_108_84ff1694c4bf92f3deef5c013ac9a5b912e6f96f_13_08_29_22_09_07_slow_weekly_tests_osx_108
http://buildlogs.mongodb.org/mci_0.9_osx_108/builds/675/test/slow_weekly_tests_0/indexbg1.js

Comment by Eric Milkie [ 30/Aug/13 ]

Yes.

Comment by Matt Kangas [ 30/Aug/13 ]

milkie, is SERVER-10505 a dup of this?

Comment by Eric Milkie [ 29/Jan/13 ]

This does not appear to be a regression from 2.2 (the test has the same issues there)

Generated at Thu Feb 08 03:17:16 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.