Work through these checks in order — most "VM won't start" or "can't connect" issues are resolved by the first few steps.

Troubleshooting Steps

1. Check the VM's status in the Azure portal. If it shows Failed rather than Stopped, click Redeploy to move it to new host hardware — this resolves many underlying host-level issues.
2. Check Boot diagnostics. On the VM's page, open Boot diagnostics to see a screenshot of the actual console — this quickly reveals issues like a Windows update stuck mid-install or a boot error message you can't otherwise see.
3. Confirm the VM is actually running, not just deallocated — a deallocated VM needs to be started, not connected to.
4. Check the Network Security Group rules for the VM and its subnet — confirm RDP (3389) or SSH (22) is actually allowed from your current IP address.
5. Confirm the OS-level firewall inside the VM isn't blocking the connection, if you're able to access it another way (e.g. via Boot diagnostics' serial console).
6. Check for a full OS disk. A Windows VM that's completely out of disk space can fail to boot properly — Boot diagnostics screenshots often reveal this directly.
7. Use the Azure Serial Console for direct text-based access to the VM's console when RDP/SSH itself is the problem, letting you run diagnostic commands without a network connection.

Common Questions

What does "Redeploy" actually do, and is it safe?

Redeploy moves the VM to new host hardware within Azure while keeping its disks and configuration intact — it's generally safe and is Microsoft's recommended first step for a VM stuck in a failed state, though it does involve a restart.

I can see the login screen in Boot diagnostics but still can't RDP in — why?

This points to a networking issue rather than the VM itself — check the NSG rules and confirm you're connecting to the correct public IP, since this is the most common cause when the VM is clearly running and responsive.