Alan Rominger
Alan Rominger
This looks like it's the same topic as https://github.com/ansible/awx/issues/6134
I know that this `update_model` method used in the cancel callback has an annoying behavior of returning `None` when a database error happens and doesn't recover. https://github.com/ansible/awx/blob/29d60844a8cee16bcd31657cecc444a692ade689/awx/main/utils/update_model.py#L43 However, I have...
I think I have seen this myself recently, but still unclear of cause. It would be good if anyone could get receptorctl information after it happens, look up the work_unit...
I've lost the logs from the case I saw, but I believe it was due to running into the postgres max connection limit.
I tried to measure how much speed boost this was compared to before (due to the deepcopy) and it was like 0.005 seconds. So nothing to get excited about.
I can no longer reproduce this issue. I have some related tests, I suggest that those be merged and this issue be closed.
This was implemented in a recent merge https://github.com/ansible/awx/pull/12875 @john-westcott-iv @hesmithrh I'm confused about what to do this issue, but it should be either Needs Test or closed.
@merickso could you give the YAML you tried to use? Here's an example of wrapping in quotes https://github.com/ansible/awx/issues/11222#issuecomment-939782609
Implemented in https://github.com/ansible/awx/pull/12744 and https://github.com/ansible/awx-operator/pull/1012
There is an open issue to support hop nodes in this context