Not able to start any type of cluster on databricks

The screenshots show two separate Azure-side capacity problems, not a Databricks configuration problem:

1. AZURE_QUOTA_EXCEEDED_EXCEPTION
   • The requested VM family quota in Central India is too low.
   • The message shows standardDCASv5Family Cores quota with Current Limit: 0 and Additional Required: 4.
   • That means clusters using that VM family cannot start until quota is available or a different VM family is used.
2. CLUSTER_LAUNCH_RESOURCE_STOCKOUT
   • The requested VM size is temporarily not available from the cloud provider in that region.
   • This is a regional capacity/SKU availability issue.

Action path:

1. Check the exact VM family or VM size configured on the cluster or pipeline.
   • If it is using DCASv5 or another constrained family, switch to a different supported instance type.
   • Community guidance in similar cases recommends trying a smaller or different VM size and checking regional availability.
2. Check Azure subscription quota for the target region.
   • In Azure Portal: Subscriptions → select the subscription → Usage + Quotas.
   • Verify the core quota for the VM family used by the cluster.
   • If the family quota is 0 or below what the cluster needs, request an increase.
3. If the workload is using serverless and the error is specifically about serverless quota, submit a quota increase request in Azure Portal:
   • Go to Help + support.
   • Select Create a support request.
   • Set Summary to quota.
   • Set Issue type to Service and subscription limits (quotas).
   • Set Quota type to Azure Databricks Serverless.
   • Enter the workspace region and the requested total DBUs/hour.
4. For classic or job clusters hitting cloud launch failures or stockout:
   • Retry later, because cloud-provider launch failures are often transient.
   • If the error mentions quota or subnet capacity, review those specifically.
   • If the same VM size keeps failing, use a different instance type or another region.
5. For pipeline cluster failures with resource exhaustion or cloud failure:
   • Restart the pipeline after changing the cluster policy or instance type.
   • If the pipeline is pinned to a specific policy or node type, update that configuration so it can use an available VM family.
6. If compute stays in Starting for a long time without a clear quota message:
   • Check the cluster Event Log for the exact failure reason.
   • If repeated retries fail across new and existing compute, open support with the cluster ID, region, and failure timestamp.

What this means for the current case:

• Increasing general vCPU quota alone is not enough if the blocked quota is for a specific VM family.
• A stockout error is not fixed by quota increase; it requires retrying later or changing VM size/region.
• The fastest workaround is usually:
  • choose a different supported VM family/size,
  • verify family quota in Usage + Quotas,
  • and retry in another region if Central India has no capacity.

References:

• Classic compute termination error codes
• Error conditions in Azure Databricks
• cluster on databricks - Microsoft Q&A
• Unable to create Databricks Cluster despite increasing total region vCores to 128 - Microsoft Q&A
• Compute Issue In databricks - Microsoft Q&A
• Serverless compute quotas