Autoscaling with unhealthy VM

Avatar
Joerg Hendricks

Hello together

Once I have one or more unhealthy VMs during a deployment/reimaging process, no further VMs will be automatically added in the autoscaling process. 
For example, I defined 50 session hosts in autoscaling, but 10 were marked as unhealthy VMs during the reimage process. As a result, only a maximum of 40 hosts are available in the host pool. Automatic expansion to 50 hosts or more does not occur.
The following options would be good, and could also be defined in the autoscale settings.
- Auto-heal/auto-delete the unhealthy VMs
- Ignore the unhealthy VM and expand the host pool back to 50 hosts.

Regards
Jörg

 

6

Comments (1 comment)

Sort by
1
Avatar
Mike Sanderson

Agree with having an option to ignore the unhealthy VM's from auto-grow. It causes us operational impact if we can't assess/remove/fix the unhealthy VM's before the auto-grow number is reached or neared and results in us having no Virtual Machines for new users to claim which is worse than having unhealthy VM's as these are unusable anyway so unsure why they are considered part of the auto-grow evaluation. 

Single Session Personal Desktop environment. 

Please sign in to leave a comment.