Matthew Jones

My feeling here is that it should remove the preferential instance group logic entirely... so no overflow or rollover, and only allows IGs for which the user has access to....

https://github.com/ansible/awx/blob/cae2c0619026ff864d24f390c3fe46b506e41f74/awx/main/access.py#L564 This is a pretty minimal set with some room to expand. So something like: * can_read ... in org... at least job template admin * can_use ... defaulted to...

This looks good to me... what's the hold up?

looking good now?

I think we could probably do this in the future with jobs directly with podman... I don't think I'd like the UX around *killing* the job specifically. Rather just setting...

This is going to be required for releasing.

No one is actively working on it. Just do the work and submit a PR

I think something is totally borked with `start` / `is-alive` / `stop`. I'll take a look

If you can make a PR for that we can get it merged in and close this out. Add an info or debug logger statement to note the skip.

Circling back to this, the only way I could see hitting this situation is if you passed ssh key data into Runner *as well as* having the file on the...