[FEATURE REQ] Adding managed RunCommand capabilities

Library name

Azure.ResourceManager.Compute

Please describe the feature.

There is an Azure Compute feature in public preview right now which allows naming of RunCommand execution and retrieval of the status of the RunCommand. https://docs.microsoft.com/en-us/azure/virtual-machines/linux/run-command-managed.

1. Currently, there is no method from VirtualMachineResource to deploy a managed run command (hitting the PUT /subscriptions/<subscriptionId>/resourceGroups/<resourceGroupName>/providers/Microsoft.Compute/virtualMachines/<vmName>/runcommands/<runCommandName> endpoint). I would like to request this feature to be added as an instance method to VirtualMachineResource.

2. The current GetVirtualMachineRunCommand and GetVirtualMachineRunCommandAsync methods can properly retrieve the status of the managed RunCommand. These methods can also expand the instanceView which contains the execution output results as long as the expand argument is equal to "instanceView". I saw this after sending my own PUT request to the endpoint I mentioned in point (1) using "TestingRunCommand" as the runCommandName and "echo 'Hello'" as the script in the request body.

   However, the plural method GetVirtualMachineRunCommands has no expand parameter and there is no GetVirtualMachineRunCommandsAsync method at all. I would like to request that the expand parameter be added and that GetVirtualMachineRunCommandsAsync be added as well.

I am in need of this feature as I am building a DevOps service for the Azure Postgres team which will deploy and retrieve the status of these managed scripts. It is necessary for the service to know whether the script execution succeeded and the output of the script execution; this data is contained in the expanded instanceView portion of the GetVirtualMachineRunCommand data. However, it is important that deploying the RunCommand and retrieiving the status can be done asynchronously in case of a server failure. Currently, if the RunCommand is executed, but the service fails there is no way of knowing whether a particular virtual machine executed the RunCommand. We are trying to keep this DevOps service stateless, so enabling these features would be very helpful.

[ArthurMa1978]

Hi @samir-puranik , yes, we will support it soon in the stable release this month.

[github-actions[bot]]

Hi @samir-puranik. Thank you for opening this issue and giving us the opportunity to assist. To help our team better understand your issue and the details of your scenario please provide a response to the question asked above or the information requested above. This will help us more accurately address your issue.