[SERVER-41235] Invariant failure !finished() src/mongo/db/concurrency/fast_map_noalloc.h 94 Created: 20/May/19 Updated: 29/Jul/19 Resolved: 29/Jul/19 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | 3.6.4 |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Minor - P4 |
| Reporter: | Christopher Watton | Assignee: | Danny Hatcher (Inactive) |
| Resolution: | Incomplete | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Running on AWS T2.medium instance OS - Amazon Linux |
||
| Operating System: | ALL |
| Participants: |
| Description |
|
We've been running these DBs for well over a couple of years now, updated to mongo 3.6.4 roughly 6 month ago. One of the members of the replica set appears to have been taken offline from the below error, we're unable to reproduce as we can't identify what request(s) occured at this time.
|
| Comments |
| Comment by Danny Hatcher (Inactive) [ 29/Jul/19 ] |
|
Closing due to lack of response. |
| Comment by Danny Hatcher (Inactive) [ 24/May/19 ] |
|
Christopher, We recommend upgrading to the latest 3.6 (3.6.12) to help avoid some concurrency issues with MMAP. Could you please do so and let us know if you experience the problem when upgraded? |
| Comment by Christopher Watton [ 23/May/19 ] |
|
Hey, it's not happened again since raising this ticket. We are deploying new DBs on WiredTiger but no current plan to migrate the existing ones |
| Comment by Danny Hatcher (Inactive) [ 22/May/19 ] |
|
It appears that there was an issue with MMAPs locking system that triggered the invariant. You mentioned that you've been using the current version for multiple months so it doesn't appear to be a common issue. Has it happened again since you've opened the ticket? Do you plan on moving to WiredTiger at some point soon? |