[SERVER-9209] Invalid access at address: 0; Got signal: 11 (Segmentation fault) Created: 02/Apr/13 Updated: 07/Apr/23 Resolved: 26/Jun/14 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | MapReduce |
| Affects Version/s: | 2.4.1, 2.4.8 |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | Robert Beekman | Assignee: | Ramon Fernandez Marina |
| Resolution: | Cannot Reproduce | Votes: | 1 |
| Labels: | core,, crash, replicaset | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Ubuntu 12.04 64bit |
||
| Operating System: | Linux |
| Participants: |
| Description |
|
Mongod crashes with a signal 11 (Segmentation fault) while running a map/reduce job. We have 2 nodes in replica with an arbiter. The first node crashed yesterday 19:30 and the second node took over, just like it should. Unfortunately the second node crashed as well at 06:35 this morning. Here are the stacktraces and errors: node1:
node2:
|
| Comments |
| Comment by Ramon Fernandez Marina [ 26/Jun/14 ] | |
|
All, since we haven't heard back for a while I'm going to mark this ticket as resolved. If this is still an issue, feel free to reopen and please provide additional information so we can try to reproduce on our end. Regards, | |
| Comment by Ramon Fernandez Marina [ 05/Jun/14 ] | |
|
krauss, I'm unable to reproduce with some sample documents. Is it possible for you to send us your dataset? If not, can you at least send us one of your documents to see how they look?
Also, the output of db.collection.stats() could be useful in trying to make progress on this issue. Regards, | |
| Comment by Kevin Krauss [ 12/Feb/14 ] | |
|
Still happening Wed Feb 12 11:21:49.327 [conn17] auth: couldn't find user yogi@yogi_berra, yogi_berra.system.users Wed Feb 12 14:52:18.262 Got signal: 11 (Segmentation fault: 11). Wed Feb 12 14:52:18.282 Backtrace: | |
| Comment by Kevin Krauss [ 19/Dec/13 ] | |
|
This is still happening on 2.4.8 on mac OSX !! please fix !! /mongodb", journal: "true", port: 27017, rest: "true", smallfiles: "true" } Thu Dec 19 14:03:35.001 Got signal: 11 (Segmentation fault: 11). Thu Dec 19 14:03:35.005 Backtrace: Here are my map reduce functions, in ruby/javascript map = %Q{ ); reduce = %Q{ ; }); map = %Q{ ); reduce = %Q{ else { result.ips[value.remote_address] = 1 }if (result.agents.hasOwnProperty(value.user_agent)) { result.agents[value.user_agent] += 1 }else { result.agents[value.user_agent] = 1 } } It will happen once every 4 out of 10 times. Also it happens when I use the mongo shell as well as the mongo ruby driver as well as mongoid which uses the moped driver. Here is my mongo.conf
also referenced in this post https://jira.mongodb.org/browse/SERVER-4441?focusedCommentId=469984&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-469984 | |
| Comment by Ruslan Sologub [ 09/May/13 ] | |
|
Hi Tad, Unfortunately I can't post the stack or log but it was doing just some regular inserts and updates, nothing special and no map/reduce. Version was 2.4.1 when crashed, now it is 2.4.3, I'll get back if this one crashes too. P.S. We use SSH tunnels to link all the replica members. Ruslan | |
| Comment by Tad Marshall [ 05/May/13 ] | |
|
Hi Ruslan, Can you post the stack trace from your crash, with a few dozen lines from before the crash for context (i.e. what was going on when the crash happened), please? Alternatively, could you attach the log file? What version of MongoDB are you running? Tad | |
| Comment by Ruslan Sologub [ 05/May/13 ] | |
|
Got this error too on our production server, but we use no map/reduce at all. We have 5 nodes in replica, two of which are hidden and usually down. The error has been thrown by a secondary member. | |
| Comment by Ben Becker [ 10/Apr/13 ] | |
|
Hi Robert, Could you attach a few additional pieces of information?
For the dump, if you could use the mongodump command with the query specifier used in the MapReduce jobs, that would be extremely helpful. If this data is sensitive, I would be happy to provide a secure SCP server you can upload to. | |
| Comment by Robert Beekman [ 02/Apr/13 ] | |
|
Just some additional info. We run the same mapreduce job with the same data on another host that runs mongo 2.2.3 and has no issues at all. |