← Back to team overview

yahoo-eng-team team mailing list archive

[Bug 1804205] [NEW] libvirtError: operation failed: domain 'instance-00000003' already exists with uuid e017717e-9647-4740-8efa-4ec2aa25f35c

 

Public bug reported:

Description
==============
Testing the Nova component, I figured out that many times the instances (servers) were not created or instantiated because the following error:

libvirtError: operation failed: domain 'instance-00000003' already
exists with uuid e017717e-9647-4740-8efa-4ec2aa25f35c

I don't understand how that is possible. There exist a way of
circumventing that? Can OpenStack solve it before creating an instance?

Worth noticing that I ran the unstack script before starting creating
new instances, so it is a residual instance from another time I ran
stack script.

Steps to Reproduce
=====================
I don't know.

Expected Result
==================
Not sure, but I think that the system needs to be prepared against this situation. The worst case is the scenario where it is not possible to create instances because a range of names is being used.

Actual Result
================
Many instances been set to error state because a name of an existent instance already exists.

Environment
==============
Devstack/Queens/Stable.
Ubuntu 16.04.

Logs & Configs
================
Attached.

** Affects: nova
     Importance: Undecided
         Status: New

-- 
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/1804205

Title:
  libvirtError: operation failed: domain 'instance-00000003' already
  exists with uuid e017717e-9647-4740-8efa-4ec2aa25f35c

Status in OpenStack Compute (nova):
  New

Bug description:
  Description
  ==============
  Testing the Nova component, I figured out that many times the instances (servers) were not created or instantiated because the following error:

  libvirtError: operation failed: domain 'instance-00000003' already
  exists with uuid e017717e-9647-4740-8efa-4ec2aa25f35c

  I don't understand how that is possible. There exist a way of
  circumventing that? Can OpenStack solve it before creating an
  instance?

  Worth noticing that I ran the unstack script before starting creating
  new instances, so it is a residual instance from another time I ran
  stack script.

  Steps to Reproduce
  =====================
  I don't know.

  Expected Result
  ==================
  Not sure, but I think that the system needs to be prepared against this situation. The worst case is the scenario where it is not possible to create instances because a range of names is being used.

  Actual Result
  ================
  Many instances been set to error state because a name of an existent instance already exists.

  Environment
  ==============
  Devstack/Queens/Stable.
  Ubuntu 16.04.

  Logs & Configs
  ================
  Attached.

To manage notifications about this bug go to:
https://bugs.launchpad.net/nova/+bug/1804205/+subscriptions


Follow ups