With hosts, there are hard broken hosts, and soft broken hosts.
A hard broken host would be if the FSLogix service fails. You want to restart it right away even if there are active sessions. If a user logs out, it will not detach the VHDX and new sessions will not work.
I would like to have one rule that controls this.
I would like to have a second rule for soft broken hosts where if the AVD agent is broken, the existing sessions will not be affected, and I can wait until they all log off before a reboot.
I don't want the same behavior where a broken AVD agent force reboots the host with active sessions like how I would want to fix the broken FSLogix service.
Autoheal Multi-config
2
I was working with a partner on some Autoscale issues and this idea came up.
Thanks for posting it!
It does make sense for the soft broken/non-user impacting issues (i.e. Needs Assistance) to have a separate set of actions to take after the users are all logged-off or during a maintenance window.
I wonder if we do it like the Pre-Stage Hosts section and allow multiple Auto-Heal Broken Hosts rules if that will work for this use case.
Or, did you have something else in mind?
Yes, this is exactly what I am thinking.
However, in my situation, I would set the FSLogix fix to happen even if users are logged in.
Also, it would be nice to have an additional option to automatically send a note to all active session users so they don't wonder why they are getting kicked out of the host.
Good call-out, Jason!
Adding a Messaging option makes a lot of sense.
We'll probably want a separate messaging option for each "Schedule" or possibly just show it if you select an "active sessions" option.
Although, I can see that quickly getting cluttered.
It might need a UI change (possibly a separate category for active sessions vs no sessions?), but I like where this idea is going. 🙂
Please sign in to leave a comment.
Comments (3 comments)