[SERVER-10292] process mongods are usually auto-disconnected, and sometimes we are fail to start it in hand Created: 23/Jul/13 Updated: 10/Dec/14 Resolved: 04/Apr/14 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Replication |
| Affects Version/s: | 2.4.4 |
| Fix Version/s: | None |
| Type: | Question | Priority: | Major - P3 |
| Reporter: | zhangjianchao | Assignee: | Unassigned |
| Resolution: | Done | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
operation system:red hat linux-server-6.4-x86_64 three Mongo config server |
||
| Attachments: |
|
| Participants: |
| Description |
|
we started three mongod(correspond to three shard) in every server.afte we insert data about 1000,0000,we check the process mongods whether they are still started yet.then we find someone is stoped, the log is uploaded in the first attachement.
how to deal with them? |
| Comments |
| Comment by Daniel Pasette (Inactive) [ 24/Jul/13 ] | ||||||||||
|
In order to diagnose how your data became corrupted we would need a good history of what happened. At this point, your only options are to start again or run repair. Once your data is in a consistent state you must monitor your import process carefully, which includes saving log files from any node impacted in a failure. | ||||||||||
| Comment by zhangjianchao [ 24/Jul/13 ] | ||||||||||
|
we just keep going to insert data, we don't know why our database is corrupted.And we test it many times, process mongods are usually auto-disconnected, we think the reason why exists query error is because the disconnected mongod. In addition, we use it in virtual machines. | ||||||||||
| Comment by Daniel Pasette (Inactive) [ 23/Jul/13 ] | ||||||||||
|
the log file you uploaded indicates that your journal files are somehow corrupted. The query error and replication error you uploaded also indicate that your database is corrupted. Do you know the steps you've taken to get the cluster to this state?
|