Short answer: “no available resources” means Azure Virtual Desktop could not find a session host with a free seat for you. Most of the time it is your own old session stuck on a host, and signing that session out from the Azure portal (or Nerdio) fixes it in a minute. If several people get it at once, the host pool is full or its hosts cannot start, and in rare cases Azure has run out of capacity for the VM size the hosts use.
An accounting firm that runs its staff desktops on Azure Virtual Desktop rang us twice in one morning. Two staff members could not log in, and both saw the same “no resources available” message while everyone else was working normally. In both cases we found an old session for that user still sitting on a host, signed it out, and they were straight back in. A few months later the same firm hit the bigger version of the problem: five staff could not get a desktop at all, because Azure had run out of capacity in Australia East for the VM size every host was built on. Here is what we found and how to fix both.
Fixed by Ryan Colombo, Brendan Ramstadius and Emma Mitchell at GCIT. Written up by Elliot Munro, CISO.
Symptoms
When you open your desktop from the Windows App, the Remote Desktop client or the web client, the connection stops and you see:
We couldn’t connect because there are currently no available resources. Try again later or if this keeps happening, ask your admin or tech support for help.
In the newer Windows App, a failed connection can also show a more general message. Select See details to see the specific reason underneath it:
An error occurred while accessing this resource. Retry the connection or contact your system administrator.

Other signs:
- One user is affected while everyone else can log in. This usually points to a stuck session for that user.
- Several users are affected at the start of the day, and the people who are already in find the desktop slow. This usually points to too few session hosts running.
- In the Azure portal, some session hosts show Shutdown or Unavailable, and the ones that are Available are at their maximum number of sessions.
- In Azure Virtual Desktop Insights or Log Analytics, the failed connections are recorded with the error
ConnectionFailedNoHealthyRdshAvailable.
People also search for this as “AVD no resources available”, “WVD no available resources”, “insufficient resources Azure Virtual Desktop” or “Windows App we couldn’t connect”.
Affected Software
- Service: Azure Virtual Desktop (pooled host pools in particular)
- Clients: Windows App, Remote Desktop client for Windows, web client, macOS and mobile clients
- Related tools: Azure portal, Nerdio Manager, autoscale and scaling plans
Cause
In a pooled host pool, each session host takes a set number of users (the maximum session limit). When you connect, Azure Virtual Desktop looks for a host that is running, healthy, allowing new sessions and not full. If it cannot find one, you get “no available resources”. There are four common reasons:
- Your old session is stuck. If your last session did not sign out cleanly, it can still be sitting on a host as a disconnected session. Disconnected and idle sessions count towards the session limit, and the broker may keep trying to send you back to that host. This was the cause in both of the single-user calls our customer made.
- Every running host is full. All active, idle and disconnected sessions take a seat. If disconnected sessions are never signed out, the pool fills up even though few people are actually working.
- Hosts are off, unhealthy or in drain mode. A host that is shut down, Unavailable, or set to not allow new sessions (drain mode) is skipped. If autoscale cannot start more hosts, the pool runs out of seats.
- Azure has no capacity for your VM size. This is rare, but it does happen. In our customer’s outage, Azure returned
AllocationFailed - We do not have sufficient capacity for the requested VM size in this regionfor the Standard_E4as_v7 size in Australia East. Stopped hosts could not start, and hosts that the automatic repair process deleted could not be rebuilt. Only 2 of 12 hosts were running, giving 8 seats for the whole firm. The subscription’s quota was nowhere near its limit, so a quota increase would not have helped.
Resolution
Step 1: Sign out the user’s old session (one user affected)
- In the Azure portal, open Azure Virtual Desktop > Host pools and select the host pool.
- Open Session hosts (or User sessions) and find the affected user’s session. It will often show as Disconnected.
- Select the session and choose Sign out (sometimes shown as Log off). If you manage the pool with Nerdio Manager, you can do the same from the host pool’s sessions list.
- Ask the user to close the client fully and connect again.
With PowerShell, an administrator can do the same using the Az.DesktopVirtualization module: find the session with Get-AzWvdUserSession and sign it out with Remove-AzWvdUserSession.
Step 2: Check the hosts (several users affected)
- In Session hosts, check the status of every host. Available is good. Shutdown and Unavailable hosts give no seats.
- Check that Allow new sessions is set to Yes on the hosts that should be taking users. A host left in drain mode after maintenance is an easy one to miss.
- Start any hosts that are shut down. If autoscale or Nerdio should have started them, check its logs for failed start jobs.
- As a stopgap, you can raise the host pool’s maximum session limit slightly so users can get onto the hosts that are running. Do this only if those hosts have the CPU and memory to cope.
Step 3: If hosts fail to start with AllocationFailed
- Check the activity log or the Nerdio job history for the start or create failure. If it says
AllocationFailedor that there is not sufficient capacity for the requested VM size, the problem is on Microsoft’s side. - Follow Microsoft’s guidance for allocation failures: try a different VM size, a different availability zone, or a different region, or use a capacity reservation for critical workloads. Raising your quota does not fix an allocation failure.
- Choose a size that is a close match. Our engineer used Standard_E4as_v6, the previous generation with the same 4 vCPUs and 32 GB of memory, because it sits on separate hardware from the v7 size that had run out.
- If the pool is managed by Nerdio or a scaling plan, make the change there, not only on the VMs. Nerdio re-applies its own host settings, so a resize made directly in Azure can be undone. In our customer’s case, resizing the existing hosts failed, so we created new hosts on the v6 size, removed the broken v7 hosts, and changed the autoscale settings so new hosts use v6.
- Test with a user from each team before you tell everyone it is fixed.
Important Notes
- Set limits for disconnected sessions. A time limit for disconnected sessions (set through Group Policy, Intune or Nerdio) signs abandoned sessions out automatically, so they stop taking seats and stop catching users on reconnect.
- Allow headroom. If the pool only just fits everyone when every host is running, one host that fails to start is enough to lock people out. Keep at least one spare host’s worth of capacity at peak times.
- Know your fallback VM size in advance. Capacity shortages are rare but last hours. Deciding ahead of time which size you would switch to, and checking your image works on it, turns an outage into a short change.
- Disconnect is not sign out. Closing the window disconnects your session but leaves it running on the host. Teach staff to use Sign out from the Start menu at the end of the day.
If Windows App itself will not open and tells you to check the Store, inspect its local package before troubleshooting AVD host availability. See Windows App: “This App Can’t Open”.
Related Resources
- Microsoft Learn: Session host statuses and health checks in Azure Virtual Desktop
- Microsoft Learn: Use drain mode to isolate session hosts
- Microsoft Learn: Configure host pool load balancing
- Microsoft Learn: Troubleshoot allocation failures when you create, restart, or resize VMs in Azure
- Fix: FSLogix “The User Profile Failed to Attach” on AVD
- Fix: “The Disk Is Full Trying to Write to C:” on RDS (FSLogix)
- Fix: “The Local Security Authority Cannot Be Contacted” in Remote Desktop
- GCIT managed IT services
- Fix: AVD Session Host Build Fails at RD Agent Install (ASR Block)
If your team relies on Azure Virtual Desktop and you want it watched and kept healthy, book a free consultation with our team.