Can an on-premise octopus server work with an octopus listening tentacle residing in azure vm?

We currently have an Octopus Server deployed on-premise (and would like to keep it that way for now) but would
like to start deploying the same applications to VMs in the cloud. If I follow the guide here ( for setting up listening tentacles. The example
shows that the new machine is added to octopus server after the 5 settings are provided: Octopus Server URL,
API Key, Environments, Roles, Listen port. Is this really going to work given that there’s no line of communication
from the VM to our on-premise Octopus Server? Or do I also need to manually add the machine in Octopus Server?
What would be the machine name in this case? Is it just the VM hostname? Thanks in advance.


Hi Tristan,

Thanks for getting in touch!

If the Tentacle VM in Azure can’t communicate back to your Octopus server, the registration command will fail (but the Tentacle will be installed and running), so you’d have to then manually add the VM from your Octopus server (as you would with a machine running on-premise) using the public IP address of the VM in Azure and port you specified.

Hope that helps!

Thank you and best regards,

Hi Henrik:
Thanks for your response. Are you referring to the registration command described on this link?
Unfortunately for me, I happen to be Azure Resource Manager instead of Azure Classic so the extensions do not appear. Instead, it appears I would have to install the tentacle like a DSC resource. Since I was in a hurry, I just chose to download the latest MSI, install it, run the scripted instance creation minus the registration and expose port 10933. I then had to add the VM from Octopus as you mentioned and it worked.

Hi Tristan,

Great to hear you got it running!

Yea, if you’re using Azure Resource Manager you have to do it all via the command line as per the PowerShell script in the section for Azure Resource Manager.

Thank you and best regards,

Hi Henrik,
Would you be able to look into another problem/question I
encountered and posted here?

What I’m basically trying to do here is have a complete build, provision,
deploy workflow driven by a checkin in
We have most of this working on-premise. However, we are trying to move to
azure and with that move, I have to ensure that the
vm is available first before I try to deploy to it and that is essentially
the first step. If the VM is not available, then I would like to
run an ARM template which creates the VM, installs all the stuff we need to
it, including the octopus tentacle. That arm template
runs without errors when we run it inside azure powershell v 1.30. My
intent is to run this in step 2 if and only if step 1 finds that the
vm does not exist. I just need to get past step 1 first :slight_smile:

Is this something you can help with?