Multiple AUTO-HEAL BROKEN HOSTS rule sets in Auto-scale

Thank you for your feature request—your input helps shape our roadmap.

Next steps:
     • We will review your request and update its status as it moves through the evaluation process.
     • If we need more details, we'll reach out in the comments.

We also welcome additional feedback and votes from the community.

I like the idea of the flexibility - would you be willing to share what you would set differently?   Maybe running a scripted action that you can customize, rather than just restarting and reimaging - when in “Needs Assistance” vs “Unavailable” ?

I've got a single pool where hosts go to “Needs assistance” status and conditions don't match Microsoft documentation on resolving that, so the workaround for now it to re-image.

With a separate auto-heal rule, I would try to reboot once, then start re-image. That would start the re-image faster than the rule for unavailable which tries to reboot 5 times.

This seems plain enough for me, you've got my vote.   A simple method to apply more than one auto-heal condition, and a unique set of actions for that particular condition that would differ from other conditions.   You've got my vote.

 

That said, I could see an additional feature to bring in here - we have scripted actions for the repair, but what about scripted actions for the condition itself?   e.g.  auto-scale runs a simple scripted action against the running hosts in a pool, and looks for a defined status code.  Then executes actions based on the code.

 

My thinking here is LOB applications, say a legacy thick client app that has some third parties plugins or services (Docuware is one like this) - the condition script could be checking to ensure it is running, and if not, an auto-heal action for that condition might be to try starting, and reboot if it can't get it started - etc.