yahoo-eng-team team mailing list archive

[Alan Baghumian <2095038@xxxxxxxxxxxxxxxxxx>]

Public bug reported:

Hello OpenStack Team!

We came across in interesting issue affecting the post deletion cleanup
of volume backed instances on Charmed OpenStack Jammy/Yoga.

After spending some time investigating, I am able to reproduce this on
my own cloud.

The general workflow is:

1. Create a volume backed instance from any image (Does not matter) -
The original issue we noticed was reported using Fedora images, but
Ubuntu images also behave the same way (--block-device
delete_on_termination=true).

2. Create an image using openstack server image create

3. Applied desired tags and mark the image as private (Optional)

4. Delete the original server

5. Launch new volume backed test servers again with --block-device
delete_on_termination=true using the image from step 2.

6. Delete the test servers

7. Check volumes and you see them all being still present.

This bug might not be related to Nova, directly but since it involves
delete_on_termination I'm filing it here for the time being.

I have a reproducer script and will attach it to the bug report below.

Please review and let me know if there are any questions.

Best,
Alan

** Affects: nova
     Importance: Undecided
         Status: New


** Tags: sts

-- 
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/2095038

Title:
  delete_on_termination=true fails to clean-up volumes in certain
  scenarios

Status in OpenStack Compute (nova):
  New

Bug description:
  Hello OpenStack Team!

  We came across in interesting issue affecting the post deletion
  cleanup of volume backed instances on Charmed OpenStack Jammy/Yoga.

  After spending some time investigating, I am able to reproduce this on
  my own cloud.

  The general workflow is:

  1. Create a volume backed instance from any image (Does not matter) -
  The original issue we noticed was reported using Fedora images, but
  Ubuntu images also behave the same way (--block-device
  delete_on_termination=true).

  2. Create an image using openstack server image create

  3. Applied desired tags and mark the image as private (Optional)

  4. Delete the original server

  5. Launch new volume backed test servers again with --block-device
  delete_on_termination=true using the image from step 2.

  6. Delete the test servers

  7. Check volumes and you see them all being still present.

  This bug might not be related to Nova, directly but since it involves
  delete_on_termination I'm filing it here for the time being.

  I have a reproducer script and will attach it to the bug report below.

  Please review and let me know if there are any questions.

  Best,
  Alan

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/2095038/+subscriptions