I’m implementing a control policy that reacts to “VM settings change” events. It tracks changes of the VM ‘annotation’ (it corresponds to the “description” field in RHEV-M), parses it, then tags the VM accordingly. The problem I’m faced with is that reading the ‘annotation’ in response to the event, I get an out-to-date value. In fact the new value is available only after a while. From measurement I made, the delay varies from about 30 s to 5 minutes.
To circumvent the problem, I delay the reading of the annotation by managing a retry in my automaton, with a 6 minutes delay. I would like to improve the automaton by looping on shorter retries (say with e.g. 30 s delays) until I detect the change. This requires to retain the value of ‘annotation’ from a retry to the next, so that to compare them and detect the change. But I’m a newbie with ManageIQ, and I don’t know how to do it.