[SERVER-32435] Automatically detect cases in which data files got removed from the OS Created: 21/Dec/17 Updated: 06/Dec/22 Resolved: 15/Aug/19 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Storage |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Improvement | Priority: | Minor - P4 |
| Reporter: | Tomer Yakir | Assignee: | Backlog - Storage Execution Team |
| Resolution: | Won't Do | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Assigned Teams: |
Storage Execution
|
| Participants: |
| Description |
|
Currently, removing data files of a running mongod process won't have an effect on the process, until it gets restarted, likely due to open file descriptors. |
| Comments |
| Comment by Eric Milkie [ 15/Aug/19 ] |
|
We aren't going to pursue this now. The difficulty of doing this outweighs the small benefit it would provide (as a user, you would need to know to set up and watch for warnings to appear in every mongod log and then act on them in a timely manner). |
| Comment by Asya Kamsky [ 18/Jan/18 ] |
|
Converted the ticket for request to log if this situation is detected. |
| Comment by Asya Kamsky [ 16/Jan/18 ] |
|
This is exactly the thing we do NOT want to do if files get deleted as those files still exist and are recoverable as long as a process exists that has file handle open on that file(s). Crashing the mongod process would make such files no longer recoverable. |