Alan Rominger

To clarify by rewording, there were 17822 jobs ran against that inventory? That is a lot, and with the volume you're working with it may not be as surprising as...

The part with ```sql UPDATE "main_jobhostsummary" SET "host_id" = NULL WHERE ... ``` would probably rule out the possibility that this is fixed by https://github.com/ansible/awx/pull/12447, since that's only dealing with...

It seems very likely that https://github.com/ansible/awx/pull/12447 could address this. I would like to get people to try that patch out.

One thing that might be useful here is that you can change some Django settings to print the SQL it runs as it migrates. That might be helpful here, because...

> It also seems weird that our historical modeling only has the relationship of hosts to ad hoc commands through the events table. Why would this seem weird? We don't...

@john-westcott-iv one last question, just to make sure. Did you test the case where you run a job, delete a host it touched, and then check that the output still...

Oh, sorry, I see that you've already documented that. Looks good.

Link https://github.com/ansible/awx/issues/7859 @ffirg we have added these workflow "prompt" fields one-by-one, and my argument is that it will take considerably less time to do a refactor where we have workflow...

Other recent prompting requests - prompting for EE (not tracked) - prompting for instance group or instance https://github.com/ansible/awx/issues/5711 It would be best to handle the workflow prompting as a separate...

Through doing https://github.com/ansible/awx/pull/12435, I've learned more, and I should apply those lessons correctly before reconsidering this. Really quick runthough: The signal callback method will only be ran in the main...