[
https://issues.apache.org/jira/browse/CLOUDSTACK-2003?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Marcus Sorensen resolved CLOUDSTACK-2003.
-----------------------------------------
Resolution: Fixed
I was actually able to trigger this on 4.0.1 as well, I misunderstood some of
the other dev's testing. I've pushed a fix to master and requested that it be
pulled into 4.0 and 4.1
> Deleting domain while deleted account is cleaning up leaves VMs expunging
> forever due to 'Failed to update resource count'
> --------------------------------------------------------------------------------------------------------------------------
>
> Key: CLOUDSTACK-2003
> URL: https://issues.apache.org/jira/browse/CLOUDSTACK-2003
> Project: CloudStack
> Issue Type: Bug
> Security Level: Public(Anyone can view this level - this is the
> default.)
> Components: Management Server
> Affects Versions: 4.0.1, 4.1.0, Future
> Reporter: Marcus Sorensen
> Priority: Blocker
> Fix For: 4.0.2, 4.1.0, Future
>
>
> Marking this as a blocker because it requires admin to manually edit database
> to fix, otherwise VMs can stack up forever in expunging under a certain
> scenario.
> Create a cloudstack zone
> Create a domain
> Create an account within the domain
> Launch VMs with the account
> Delete the account, this should begin stopping/destroying VMs associated with
> the account
> Before the account job is finished, delete the domain. Now Domain is gone and
> account is gone, but associated VMs never clean up.
> Expunging will error out due to the domain missing:
> 2013-04-10 16:34:52,925 ERROR [cloud.resourcelimit.ResourceLimitManagerImpl]
> (Job-Executor-7:job-39) Failed to update resource count for account
> 2013-04-10 16:34:53,019 WARN [cloud.vm.UserVmManagerImpl]
> (Job-Executor-7:job-39) Concurrent operations on expunging
> VM[User|7646b383-7cd7-48a1-a056-a48fb54111c8]
> com.cloud.exception.ConcurrentOperationException: Failed to transit state
> at
> com.cloud.storage.StorageManagerImpl.destroyVolume(StorageManagerImpl.java:2250)
> at
> com.cloud.utils.component.ComponentInstantiationPostProcessor$InterceptorDispatcher.intercept(ComponentInstantiationPostProcessor.java:125)
> at
> com.cloud.storage.StorageManagerImpl.cleanupVolumes(StorageManagerImpl.java:3725)
> at
> com.cloud.utils.component.ComponentInstantiationPostProcessor$InterceptorDispatcher.intercept(ComponentInstantiationPostProcessor.java:125)
> at
> com.cloud.vm.VirtualMachineManagerImpl.advanceExpunge(VirtualMachineManagerImpl.java:409)
> at com.cloud.vm.UserVmManagerImpl.expunge(UserVmManagerImpl.java:1776)
> at
> com.cloud.user.AccountManagerImpl.cleanupAccount(AccountManagerImpl.java:578)
> at
> com.cloud.user.AccountManagerImpl.deleteAccount(AccountManagerImpl.java:515)
> at
> com.cloud.user.AccountManagerImpl.deleteUserAccount(AccountManagerImpl.java:1180)
> If I manually re-enable the domain in the database, expunging completes. Then
> I can safely delete the account. This is a regression from the 4.0 behavior,
> according to our test suites, where account and domain could be safely
> removed and associated resources would clean up.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
