← Back to team overview

yahoo-eng-team team mailing list archive

[Bug 1938925] [NEW] State not reset during nova-compute service init

 

Public bug reported:

Description
===========
Some transient state are cleaned during the nova-compute service init.
We encountered an issue with other states during nova backup action and shelve action.

Steps to reproduce
==================
nova backup
1/ Start Backup on the instance
2/ Restart nova-compute
3/ Instance stuck in image_backup 

shelve
1/ Shelve an instance
2/ restart nova-compute during state shelving / shelving_image_pending_upload / shelving_image_uploading
3/ instance stuck in the previous state

Expected result
===============
Nova can't manager the job exec by libvirt because after a restart a lost the context so the jobs must be abort during the nova-compute restart

Actual result
=============
The instance is stuck in a transient state, user is blocked

** Affects: nova
     Importance: Undecided
     Assignee: Pierre Libeau (pierre-libeau)
         Status: New

** Changed in: nova
     Assignee: (unassigned) => Pierre Libeau (pierre-libeau)

-- 
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/1938925

Title:
  State not reset during nova-compute service init

Status in OpenStack Compute (nova):
  New

Bug description:
  Description
  ===========
  Some transient state are cleaned during the nova-compute service init.
  We encountered an issue with other states during nova backup action and shelve action.

  Steps to reproduce
  ==================
  nova backup
  1/ Start Backup on the instance
  2/ Restart nova-compute
  3/ Instance stuck in image_backup 

  shelve
  1/ Shelve an instance
  2/ restart nova-compute during state shelving / shelving_image_pending_upload / shelving_image_uploading
  3/ instance stuck in the previous state

  Expected result
  ===============
  Nova can't manager the job exec by libvirt because after a restart a lost the context so the jobs must be abort during the nova-compute restart

  Actual result
  =============
  The instance is stuck in a transient state, user is blocked

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/1938925/+subscriptions