[SERVER-43247] crash sigsegv Created: 10/Sep/19 Updated: 11/Nov/19 Resolved: 11/Nov/19 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | 4.0.12 |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | krzysztof osmulski | Assignee: | Danny Hatcher (Inactive) |
| Resolution: | Done | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Attachments: |
|
| Participants: |
| Description |
|
i play around with mongo a bit. standalone. I got this crash recently. —
— |
| Comments |
| Comment by krzysztof osmulski [ 10/Nov/19 ] | ||||||||||||||||||||||||||||||||||||||||||||||||||
|
Yes | ||||||||||||||||||||||||||||||||||||||||||||||||||
| Comment by Danny Hatcher (Inactive) [ 08/Nov/19 ] | ||||||||||||||||||||||||||||||||||||||||||||||||||
|
Were you ever able to successfully run the --repair? | ||||||||||||||||||||||||||||||||||||||||||||||||||
| Comment by Danny Hatcher (Inactive) [ 20/Sep/19 ] | ||||||||||||||||||||||||||||||||||||||||||||||||||
|
I recommend specifying the same logpath with the --repair argument that you do when you normally launch a mongod process. That way the logs will be consistent and you can provide them easily. | ||||||||||||||||||||||||||||||||||||||||||||||||||
| Comment by krzysztof osmulski [ 16/Sep/19 ] | ||||||||||||||||||||||||||||||||||||||||||||||||||
|
Hello. I simply cannot find it. Mean the 'verify' logs comming from --repair. So i see two options.
| ||||||||||||||||||||||||||||||||||||||||||||||||||
| Comment by Danny Hatcher (Inactive) [ 16/Sep/19 ] | ||||||||||||||||||||||||||||||||||||||||||||||||||
|
Can you please provide the full mongod log covering the last --repair attempt up to the crash? | ||||||||||||||||||||||||||||||||||||||||||||||||||
| Comment by krzysztof osmulski [ 14/Sep/19 ] | ||||||||||||||||||||||||||||||||||||||||||||||||||
|
Yes, I know i got corrupted data but could not find a reason to it. I now did mongod --repair succesfully and today got back the: — 2019-09-14T06:00:10.822+0200 I NETWORK [conn4265] received client metadata from 127.0.0.1:33518 conn4265: { driver: { name: "mongo-java-driver", version: "3.9.1" }, os: { type: "Linux", name: "Linux", architecture: "amd64", version: "4.4.0-148-generic" }, platform: "Java/Oracle Corporation/1.8.0_201-b09" } —
The server did not crash not restart in period from repair to now. What more i can blame is the SSD but the S.M.A.R.T show good health. Besides the sigsegv what is raised here is there any checks i could possibly do to isolate issue? Form me it seems that for some reason data got corrupted by multiconnection access that search and update at the same time.
The slave is not an option for now. This is not that level, i use mongo for small analytics purposes. But it seems that it simply broke itself upon heavier load | ||||||||||||||||||||||||||||||||||||||||||||||||||
| Comment by Danny Hatcher (Inactive) [ 12/Sep/19 ] | ||||||||||||||||||||||||||||||||||||||||||||||||||
|
It appears that you experienced corruption in the underlying data files; this is likely related to the underlying infrastructure of the server failing. We recommend using Replication to spread multiple mongod process across servers to easily recover from issues like this. I see that you ran the repairDatabase command but it did not succeed. This command was actually deprecated in 4.2 as it does not cover as many cases as the --repair configuration option. Please try starting the server with the --repair option and let it run. When the repair process is finished, please try restarting the node without that option and see if you experience any problems. | ||||||||||||||||||||||||||||||||||||||||||||||||||
| Comment by krzysztof osmulski [ 10/Sep/19 ] | ||||||||||||||||||||||||||||||||||||||||||||||||||
|
te server is standalone running on ext4, ubuntu 14.04
— process was around 3GB including 2GB specified mongo cache. Now to make it more complicated in attached mongo.log you will find more crashes (this bug is last one). I hope You get something meaningful from that output what will make mongo reliable as standalone for big collections | ||||||||||||||||||||||||||||||||||||||||||||||||||
| Comment by Danny Hatcher (Inactive) [ 10/Sep/19 ] | ||||||||||||||||||||||||||||||||||||||||||||||||||
|
In order for us to diagnose this problem, can you please describe the situation when you encountered the stacktrace? Were you running a series of specific queries? Did you just shutdown and then bring the server back up again? Did the hardware underneath the process experience a failure? Can you please also provide the full mongod log covering a significant time period before the crash up to and including the crash itself? |