For more than 8 hours we are experiencing connectivity issues between Worker Role (.NET bundle) & Azure VM (RabbitMQ server on Windows Server). The code was not changed - it was working without any issues like that for more than 1 month (even yesterday during the highload period of our system everything was OK).
We've got the following issues
- Failed to receive confirmations from Rabbitmq even for the smallest messages (WaitForConfirmsOrDie exceptions)
- During the fault periods - Azure VM is unaccessable via RDP & Rabbitmq management port
- We are not seeing any high CPU usage on both VM & Cloud service
- We've restarted Azure VM & we have lost RDP connectivity to it (connection drops after password & certificate verify dialogs)
We had to redeploy the Stage environment for several times - and each try was stable for only 2-3 minutes. We can see the issues with connectivity even from the external network (using our office&home PCs).