Failed to Migrate Stopped Instance

LXD
1

Hi, when I try to migrate a stopped instance, I got this error.
Failed to migrate ceph RBD volume: Failed to run: rbd --id admin --cluster ceph --pool lxd unmap container_shpc-153-instance-E11blLGm: exit status 16 (rbd: sysfs write failed rbd: unmap failed: (16) Device or resource busy)

I sure that the instnace is stopped. LXD also tell me that the instance is stopped.
Snipaste_2023-05-14_09-35-53

I found that some client keep active between ceph, but I can not foud out which the concrete client process. What can I do to avoid this error?

Snipaste_2023-05-14_09-36-05
Snipaste_2023-05-14_09-36-05782×405 369 KB

2

I got this error after upgrading lxd and kernel version. Now my lxd version is 5.12-c63881f and kernel version is Linux 5.15.0-71-generic x86_64 (ubuntu 20.04).

Any method to debug this issue? :face_holding_back_tears:

3

Please can you check this on a supported version of LXD, which at the current time is LXD 5.13 for the latest/stable channel.

Thanks

4

Is this a known problem in 5.12?

5

No, but we did a lot of rework and refactoring on the migration subsystem in LXD 5.13 so would need a reproducer on that before we could look at it anyway.

6

ok, thank you. I try upgrade our lxd version