-
Notifications
You must be signed in to change notification settings - Fork 165
[1LP][RFR] Update snapshot tests to use create_vm fixture #9942
Conversation
…ut_description tests to use create_vm
PRT is invalid for some reason. I am not sure why. I have initiated a rerun. I tested each test function in test_snpshot.py locally and all passed except for test_verify_revert_snapshot, as I mentioned in earlier comment. I'll be fixing this soon. |
PRT keeps going to Invalid. Might be an issue with it. I'll bring it up Monday. No need to bother anyone on a Sunday. |
I detected some fixture changes in commit f3f27dd The local fixture
Please, consider creating a PRT run to make sure your fixture changes do not break existing usage 😃 |
I updated create_vm by wrapping up vm.cleanup_on_provider in a wait_for with HandleException=True. I can add exception handling directly to cleanup_on_provider in another PR or do it here? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Just one question about your change to the local create_vm
in test_vm_clone.py
. Changes LGTM though and PRT results show that the vm-name has your nic in it. PRT failures are not a result of changes here.
I added a wait_for to added a wait_for to cleanup_on_provide that has handle_exception=True by default. It was suggested we use kwarg to control if we want to handle exceptions in the cleanup, but I don't know how this get used as we do not call create_vm directly (it's a fixture). |
@prichard77 Please check PRT failure |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nothing much going on in this PR, mostly just s/name/othername/g
Good job anyway. It seems like the create_vm
gets some momentum.
I added a wait_for to added a wait_for to cleanup_on_provide that has handle_exception=True by default. It was suggested we use kwarg to control if we want to handle exceptions in the cleanup, but I don't know how this get used as we do not call create_vm directly (it's a fixture). |
cfme/common/vm.py
Outdated
"""Clean up entity on the provider if it has been created on the provider | ||
|
||
Helper method to avoid NotFoundError's during test case tear down. | ||
""" | ||
if self.exists_on_provider: | ||
self.mgmt.cleanup() | ||
wait_for(lambda: self.mgmt.cleanup, handle_exception=handle_cleanup_exception, | ||
timeout=900) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is a lengthy timeout, if cleanup fails for some reason and we are stuck because handle_exception=True
, we could spend 15 min here. I think 300 or maybe even 180 would be fine here.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Reduced to 300.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
1 question about timeout, but LGTM, moving to [1LP][WIPTEST]
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good! waiting on evaluation of PRT failures
I mentioned in earlier comments that AFAIK the failures are for tests with BZs against them and/or were failing before the change was made to use create_vm. |
Thanks PJ! |
What are we waiting on here? i have added several comments on the PRT results and we have 3 approvals. Did I miss something? |
@prichard77 it was 1LPWIPTEST during freeze on Friday, and was marked 1LPRFR at 5:24PM EST. Nothing gets merged until the release is tagged, which it was last night. 2nd level reviewers haven't come back to the RFR lists to re-review yet. |
Purpose or Intent
PRT Run
The function test_verify_revert_snapshot fails for rhv providers. This was failing before the update to create_vm and is not caused by it. I am going to begin work on fixing test_verify_revert_snapshot for rhv providers in another PR shortly.