summary refs log tree commit diff stats
path: root/results/classifier/gemma3:12b/hypervisor/975
blob: e95f228540f21bf54995bdb34d908c806cf013a5 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
LXD with QEMU 6.2.0 (and 7.0.0-rc3) breaks during stateful migration
Description of problem:

Steps to reproduce:
```
sudo snap install --lxd
sudo lxd init --auto
lxc init images:ubuntu/20.04/cloud v1 --vm
Creating v1
lxc config device override v1 root size.state=2GiB
Device root overridden for v1
lxc config set v1 migration.stateful=true
lxc start v1
sleep 10
lxc exec v1 -- uptime
 22:05:54 up 0 min,  0 users,  load average: 0.07, 0.02, 0.00
lxc snapshot v1 --stateful
Error: Migration call failed
lxc snapshot v1 --stateful
Error: Monitor is disconnected
```

The first attempt at `lxc snapshot v1 --stateful` caused this in the `lxc info v1 --show-log` log output:

```
qemu-system-x86_64: qemu_savevm_state_complete_precopy_non_iterable: bdrv_inactivate_all() failed (-1)
```

The second attempt caused this:

```
qemu-system-x86_64: ../block.c:6757: bdrv_inactivate_recurse: Assertion `!(bs->open_flags & BDRV_O_INACTIVE)' failed.
```

Which crashed QEMU completely and caused the VM to die.
Nothing relevant showed up in dmesg, so this wasn't caused by an obvious seccomp or apparmor policy issue.
Additional information:
Originally reported by Stephane Graber at https://github.com/lxc/lxd/issues/9875