We've recently begun experiencing capacity constraints on our preferred AVD VM size. Although I've configured alternate sizes in Azure Capacity Extender for the affected pools (which works great in most scenarios), I've noticed that some hosts fail to start for an Application Management task due to capacity issues. Therefore, I'd like to suggest one or more of the following solutions to this issue:
- Chain Application Management to the reimage task, so that the application deployments can be completed without stopping and restarting the VM after reimaging or between application deployments.
- Allow Application Management for stopped VMs to be delayed and queued until those VMs are started by Auto-scale working hours settings.
- Apply the Azure Capacity Extender logic to resize the VM when it is started for Application Management.
- Trigger an Application Management task that has failed for this reason to resume when the VM is started. Currently, the affected Application Management tasks cannot even be manually resumed, resulting in VMs missing security agents for the day, applications needing to be deployed manually, or VMs being swapped out.
Comments (0 comments)