We have an AVD deployment but we have limited control over it. Another team owns the tenant and has created a subscription for our team. We are responsible for creating our image, but they manage the creation of host pools and deploying the image even though they are deployed to our subscription (It's a weird, difficult arrangement). We have no insight into how this is done and likely never will.
Lately we've been seeing an issue with some of our host pools that has prevented users from logging into their session hosts. The session host will start and be marked as running, but it remains unavailable. This doesn't happen to all of the session hosts in a host pool, e.g. I can log into my session host, but others in the same host pool are no longer able to log into theirs. They could until recently.
My suspicion is that the issue is due to the image that was deployed. I used Win10 Ent N as the base for our image without knowing it is not supported in AVD. The team doing the deployment had to take manual steps to convert the image to a supported version (there's a reason I couldn't just spin up a new version, but that isn't relevant). I don't know exactly what those steps were but the phrase "cloned to a supported version" has been used. I imagine whatever they did left us with a less than optimal deployment, though.
Their explanation is that the problem is due to the age of the host pools. I'm skeptical of this explanation because we have an ongoing network issue that they also attributed to host pool age without doing any troubleshooting. They backtracked when we provided evidence that the issue was affecting VMs in new host pools, VMs in old host pools, and VMs that aren't in any host pools. And while I haven't seen much about it, what I have seen implies that only the session hosts need to be replaced while the host pools can be used indefinitely.
Are my assumptions correct that host pools don't need to be replaced whenever we deploy a new image? Is the other team correct that they do degrade over time?
