[SERVER-496] ASSERT failure BtreeBucket Created: 18/Dec/09 Updated: 17/May/10 Resolved: 17/May/10 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | 1.1.3 |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Minor - P4 |
| Reporter: | Doug Green | Assignee: | Eliot Horowitz (Inactive) |
| Resolution: | Cannot Reproduce | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
unix with mongo master/slave, this happened on the master |
||
| Attachments: |
|
| Participants: |
| Description |
|
We killed mongodb and the error stopped. So far it looks like no data corruption. And it looks like the master-slave replication stayed sane. I've marked this as priority:minor because I don't expect to run this script again. But as a mongod server failure, you might set a higher priority. I ran a cleanup script to fix bad data caused by previous bugs in our software. The script queried from a large collection (url) on our slave, then started updating records on our master. We we're updating using a key. The script got killed and was restarted. As far as I can tell, the original script was dead, but based on the symptoms here, I'll guess that we had two versions of this script running, and it was the "killed" script that was causing these problems. (That's the only explanation I have for these updates showing up in our log after the script was apparently done). The script was basically running this update (I tried to translate from php to json, so there may be syntax errors below, hopefully you get the idea). We ran this on about 10,000 scan_id's, each matching many records:. mongo> db.url.update( {scan_id' : 12345}, {'$set' : {'scan': {client_id:9876, active:0, status:0}}}, {multiple:true}) The collection is big: mongo> db.url.stats() And this is what we saw in our logs: Fri Dec 18 18:41:23 connection accepted from 192.168.1.200:39880 #2348670 Fri Dec 18 18:41:23 scan_stats.url Assertion failure false db/btree.cpp 713 exception assertion db/btree.cpp:713 6453ms |
| Comments |
| Comment by Eliot Horowitz (Inactive) [ 17/May/10 ] |
|
Can't reproduce at some point, lots of fixes since 1.1.3 |
| Comment by Doug Green [ 18/Dec/09 ] |
|
log file starting from about an hour before problem, up until the reboot |
| Comment by Doug Green [ 18/Dec/09 ] |
|
I don't know how long these pastie's stay around, but here are the pastie details I posted in IRC: |