[SERVER-1024] Segmentation fault while building indexes Created: 15/Apr/10  Updated: 17/May/10  Resolved: 17/May/10

Status: Closed
Project: Core Server
Component/s: Index Maintenance
Affects Version/s: 1.4.0
Fix Version/s: None

Type: Bug Priority: Major - P3
Reporter: Jeremy Andrews Assignee: Eliot Horowitz (Inactive)
Resolution: Cannot Reproduce Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

64-bit Centos Linux.


Participants:

 Description   

While building indexes, MongoDB core dumped.

Here are the relevant lines from the log, please let me know if you need anything else.

Thu Apr 15 15:57:36 building new index on

{ @bundle: 1, @fields.path_fragment.value: 1 }

for examiner_stage.fields_current.taxonomy_term
Thu Apr 15 15:57:36 Buildindex examiner_stage.fields_current.taxonomy_term idxNo:7 { _id: ObjId(4bc798b98ead0e2a781e0000), ns: "examiner_stage.fields_current.taxonomy_term", key:

{ @bundle: 1, @fields.path_fragment.value: 1 }

, name: "term_by_path" }
Thu Apr 15 15:57:38 insert examiner_stage.system.indexes 1782ms
Thu Apr 15 15:57:38 Got signal: 11 (Segmentation fault).
Thu Apr 15 15:57:38 Backtrace:
0x6a7ea9 0x37784302d0 0x5b6739 0x5b6a9f 0x5b6cfe 0x560332 0x5fc9ad 0x5feb83 0x6a9a72 0x6bbde0 0x3198a06617 0x37784d3c2d
/usr/local/mongodb/bin/mongod(_ZN5mongo10abruptQuitEi+0x399) [0x6a7ea9]
/lib64/libc.so.6 [0x37784302d0]
/usr/local/mongodb/bin/mongod(_ZN5mongo11BtreeCursor4initEv+0x29) [0x5b6739]
/usr/local/mongodb/bin/mongod(_ZN5mongo11BtreeCursor12initIntervalEv+0x7f) [0x5b6a9f]
/usr/local/mongodb/bin/mongod(_ZN5mongo11BtreeCursor7advanceEv+0x14e) [0x5b6cfe]
/usr/local/mongodb/bin/mongod(_ZN5mongo7getMoreEPKcixRNS_5CurOpE+0x352) [0x560332]
/usr/local/mongodb/bin/mongod(_ZN5mongo15receivedGetMoreERNS_10DbResponseERNS_7MessageERNS_5CurOpE+0x25d) [0x5fc9ad]
/usr/local/mongodb/bin/mongod(_ZN5mongo16assembleResponseERNS_7MessageERNS_10DbResponseERK11sockaddr_in+0x11e3) [0x5feb83]
/usr/local/mongodb/bin/mongod(_ZN5mongo10connThreadEv+0x242) [0x6a9a72]
/usr/local/mongodb/bin/mongod(thread_proxy+0x80) [0x6bbde0]
/lib64/libpthread.so.0 [0x3198a06617]
/lib64/libc.so.6(clone+0x6d) [0x37784d3c2d]
Thu Apr 15 15:57:38 dbexit:
Thu Apr 15 15:57:38 shutdown: going to close listening sockets...
Thu Apr 15 15:57:38 going to close listening socket: 10
Thu Apr 15 15:57:38 going to close listening socket: 12
Thu Apr 15 15:57:38 shutdown: going to flush oplog...
Thu Apr 15 15:57:38 shutdown: going to close sockets...
Thu Apr 15 15:57:38 shutdown: waiting for fs preallocator...
Thu Apr 15 15:57:38 shutdown: closing all files...
Thu Apr 15 15:57:38 got request after shutdown()
41/45 91%
Thu Apr 15 15:57:40 closeAllFiles() finished
Thu Apr 15 15:57:40 shutdown: removing fs lock...
Thu Apr 15 15:57:40 dbexit: really exiting now
ERROR: Client::~Client _context should be NULL: conn



 Comments   
Comment by Jeremy Andrews [ 16/Apr/10 ]

I restored this same backup on 3 different servers and then ran the same script to create indexes on each, and only experienced the above segfault the one time. All instances are running fine at this time. I am therefore unable to duplicate this bug. If the above debug isn't enough to be helpful, feel free to close this bug report.

Comment by Jeremy Andrews [ 15/Apr/10 ]

I restarted mongod and then re-ran the script that creates our indexes, and it worked perfectly the second time. The database has not had any problems since the above was reported.

Comment by Eliot Horowitz (Inactive) [ 15/Apr/10 ]

is it possible to attach the raw data files?
can move to support so its private if needed

Generated at Thu Feb 08 02:55:50 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.