We have a new databricks instance on Azure and looks like capacity will not be enough. So my question is if we create a new subnet with larger capacity can we just replace the existing subnet on the current instance of databricks with a new subnet? And if we can replace the existing subnet, then how can we go about it?
Or is it possible to add an additional subnet to azure databricks instance?
This is not possible right now. The solution would be to create new workspace and migrate everything into it. The Databricks Labs project has two projects: migrate and databricks-sync that could be used for migration between workspaces.
Update, October 2023: databricks-sync isn't maintained anymore, and migrate has the coverage for a limited set of resources. Current recommendation is to use Databricks Terraform Exporter that supports broader set of resources. Although it doesn't support Hive Metastore export - the corresponding functionality could be used from the migrate tool.