← Back to team overview

yahoo-eng-team team mailing list archive

[Bug 2111558] [NEW] permission denied when dong hard-reboot on instance with snapshot

 

Public bug reported:

Test environment : OpenStack Caracal 2024.1 (installed with OpenStack-
ansible, tag 29.2.2) running on Ubuntu 22.04 (slurp update coming from
Antelope 2023.1).

What we experience (all done via horizon) :
- create a volume (hosted on an NFS-backend)
- create an instance of this volume
- instance is pingable and accessible with ssh
- do a hard reboot in horizon
- instance reboots and is again in working order
- now create a snapshot of this instance
- do again a hard reboot
- horizon throws a libvirt error, no xml file is created in /etc/libvirt/qemu, instance goes to error

- logs : : libvirt.libvirtError: internal error: process exited while
connecting to monitor: 2025-05-23T06:47:27.126623Z qemu-system-x86_64:
-blockdev
{"driver":"file","filename":"/var/lib/nova/mnt/081516f3de9900246707a27a6db4bda9/volume-03a81380-8ce4-4570-ba7c-7057eb898460","aio":"native","node-
name":"libvirt-2-storage","cache":{"direct":true,"no-
flush":false},"auto-read-only":true,"discard":"unmap"}: Could not open
'/var/lib/nova/mnt/081516f3de9900246707a27a6db4bda9/volume-03a81380-8ce4-4570-ba7c-7057eb898460':
Permission denied

So, the original volume becomes the backing file but libvirt gets an
error on this backing file : permission denied. This does not happen
with a soft-reboot or a live-migration.

Further operations
- delete the instance, delete the snapshot
- the original volume is again usable to create an instance

thx
Peter

** Affects: nova
     Importance: Undecided
         Status: New

** Description changed:

  Test environment : OpenStack Caracal 2024.1 (installed with OpenStack-
  ansible, tag 29.2.2) running on Ubuntu 22.04 (slurp update coming from
  Antelope 2023.1).
  
  What we experience (all done via horizon) :
  - create a volume (hosted on an NFS-backend)
  - create an instance of this volume
  - instance is pingable and accessible with ssh
  - do a hard reboot in horizon
- - instance reboots and is again in working order 
+ - instance reboots and is again in working order
  - now create a snapshot of this instance
  - do again a hard reboot
  - horizon throws a libvirt error, no xml file is created in /etc/libvirt/qemu, instance goes to error
  
  - logs : : libvirt.libvirtError: internal error: process exited while
  connecting to monitor: 2025-05-23T06:47:27.126623Z qemu-system-x86_64:
  -blockdev
  {"driver":"file","filename":"/var/lib/nova/mnt/081516f3de9900246707a27a6db4bda9/volume-03a81380-8ce4-4570-ba7c-7057eb898460","aio":"native","node-
  name":"libvirt-2-storage","cache":{"direct":true,"no-
  flush":false},"auto-read-only":true,"discard":"unmap"}: Could not open
  '/var/lib/nova/mnt/081516f3de9900246707a27a6db4bda9/volume-03a81380-8ce4-4570-ba7c-7057eb898460':
  Permission denied
  
  So, the original volume becomes the backing file but libvirt gets an
  error on this backing file : permission denied. This does not happen
- with a soft-reboot, only with operations that re-create the libvirt xml
- file like hard-reboot and live-migration.
+ with a soft-reboot or a live-migration.
  
  Further operations
  - delete the instance, delete the snapshot
  - the original volume is again usable to create an instance
  
- 
  thx
  Peter

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to OpenStack Compute (nova).
https://bugs.launchpad.net/bugs/2111558

Title:
  permission denied when dong hard-reboot on instance with snapshot

Status in OpenStack Compute (nova):
  New

Bug description:
  Test environment : OpenStack Caracal 2024.1 (installed with OpenStack-
  ansible, tag 29.2.2) running on Ubuntu 22.04 (slurp update coming from
  Antelope 2023.1).

  What we experience (all done via horizon) :
  - create a volume (hosted on an NFS-backend)
  - create an instance of this volume
  - instance is pingable and accessible with ssh
  - do a hard reboot in horizon
  - instance reboots and is again in working order
  - now create a snapshot of this instance
  - do again a hard reboot
  - horizon throws a libvirt error, no xml file is created in /etc/libvirt/qemu, instance goes to error

  - logs : : libvirt.libvirtError: internal error: process exited while
  connecting to monitor: 2025-05-23T06:47:27.126623Z qemu-system-x86_64:
  -blockdev
  {"driver":"file","filename":"/var/lib/nova/mnt/081516f3de9900246707a27a6db4bda9/volume-03a81380-8ce4-4570-ba7c-7057eb898460","aio":"native","node-
  name":"libvirt-2-storage","cache":{"direct":true,"no-
  flush":false},"auto-read-only":true,"discard":"unmap"}: Could not open
  '/var/lib/nova/mnt/081516f3de9900246707a27a6db4bda9/volume-03a81380-8ce4-4570-ba7c-7057eb898460':
  Permission denied

  So, the original volume becomes the backing file but libvirt gets an
  error on this backing file : permission denied. This does not happen
  with a soft-reboot or a live-migration.

  Further operations
  - delete the instance, delete the snapshot
  - the original volume is again usable to create an instance

  thx
  Peter

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/2111558/+subscriptions