This is an example template to deploy an Azure virtual machine running Polynote.

The simplest way to deploy is to just click the "Deploy to Azure" button here. Read on if you want to do it from the command line instead.

And you end up with the following ready to go

The following commands can be used to deploy the template from the command line using the Azure CLI.

# Create a resource group to deploy the virtual machine to
az group create -g <resource group name> -l <location name>

# For example
az group create -g polynote -l northeurope

# Deploy the template
az deployment group create \
    -g <resource group name> \
    --template-file azuredeploy.bicep \
    --parameters resourcePrefix='tstvm' clientIpAddress='<your ip address>' vmAdminUser='<admin username>' vmAdminPass='<admin password>' \
    --verbose

As Polynote currently does not have any built-in security, the template creates a network security group blocking access except for a given IP address on ports 22 (SSH) and 8192 (Polynote). You can find your IP address using a service such as ipify.

By default the template will deploy a Standard_D2s_V3 server running Ubuntu 18.04 LTS. The size of the VM can be changed to any of those available in your region which can be found using the following Azure CLI command.

az vm list-sizes -o table

When the deployment is complete the polynote service should be started, so you can point your browser to http://<vm name>.<region>.cloudapp.azure.com:8192 and start playing. You can also SSH onto the virtual machine and have a look around.

N.B. The default port for Polynote is 8192, if you want to change this you will need to modify the YAML configuration file as per the documentation.

You can find the DNS name for your VM either via the Azure Portal or from the command line.

# Get the VM details and filter to the name and fully qualified DNS
az vm list -g <resource group name> -d --query "[].{name: name, fqdns: fqdns}" --output table

If you want to filter this list based on number of cores or memory available then you can do this using the --query parameter which utilizes JMESPath. Remember to change the location to the Azure region you want to deploy to.

N.B. All commands are outputting in a table format for ease of reading.

You can list the locations available to you as an option using the Azure CLI as follows.

az account list-locations --output table

For example, to find all DS_v3 series sizes with more than 8Gb of memory in North Europe you could use the following query.

az vm list-sizes --location northeurope --query "[?starts_with(name, 'Standard_D') && ends_with(name, 's_v3') && memoryInMb > `8192`]" -o table

REMEMBER if you're using Powershell to query using the Azure CLI you will need to escape the back ticks (see the memory part of the query below).

az vm list-sizes --location northeurope --query "[?starts_with(name, 'Standard_D') && ends_with(name, 's_v3') && memoryInMb > ``8192``]" -o table

Based on my own subscription this outputs the following (at the time of writing):