Red Hat Bugzilla – Bug 1256826
[RFE] Alter generation of fail-safe snapshot delete so that it is created before snapshot request is presented to provider.
Last modified: 2018-01-04 13:26:26 EST
Description of problem: BZ 1256821 reports issue where snapshot request to openstack results in a reported connection error, but the snapshot is created, never to be deleted, consuming disk space in provider storage environment.
Version-Release number of selected component (if applicable): 220.127.116.11
How reproducible: refer to BZ 1256821
Steps to Reproduce:
the fail-safe snapshot delete was introduced into CFME as a means to ensure that regardless of what happened to the vmscan processing, any snapshot file created in the provider environment would be removed after not more than 60 minutes. The BZ 1256821 demonstrates that generating that snapshot fail-safe request *after* the snapshot create request is not sufficient to achieve the desired results, (ie, do no harm to the provider storage environment). Createing the failsafe request before requesting the snapshot create will preserve as best we can the attempt to return the storage environment back to the state before the snapshot was created.