I'm facing an issue with my Azure Databricks instance that has been stuck in the deleting status for more than 48 hours. I've attempted to forcefully delete it using both CLI and PowerShell, but nothing seems to work. There are no locks on any resources, and I continue to incur charges for this workspace. I also have a developer support plan, but I've been advised to submit a ticket, which apparently requires me to upgrade to a standard support plan. Do I really need to upgrade just to create a support ticket for this issue? Is there any other way to resolve this?
3 Answers
Is your setup able to reach the Azure management plane via DNS? Sometimes network settings can impact the deletion process if there’s an issue there.
When your Databricks workspace is stuck in a deleting state, it often indicates that Azure is still managing some resources in the background. Check the managed resource group that Databricks created for any leftover resources; sometimes they need to be cleared out first before the deletion can finish. If anything is still active in that group, it might be what's holding things up.
I've noticed similar issues before. Everything in the managed group is still there, but since it's a managed resource group, you can't delete it manually.
Unfortunately, once a cloud resource hangs in the deleting status, it's typically a backend issue that only Azure support can fix. If all your attempts to delete it have failed and there are no resource locks, then it’s likely stuck in Azure’s control plane. Some folks manage to get help by reaching out on Azure forums or tagging Microsoft engineers on social media, but it can be hit or miss. Considering the ongoing charges, you may have no choice but to upgrade your support plan if you want a quick resolution.
I’m pretty much leaning towards upgrading my plan today to get this sorted. It's frustrating to watch the costs pile up!
Yes, everything's fine on that end; the vNet can reach the management plane, and there are no user-defined routes affecting connectivity.