This typically points to delayed compute provisioning behind the SQL Warehouse, often due to temporary capacity/resource availability or a transient startup issue.

A few things I would try:

1. Stop and restart the SQL Warehouse
If it has been stuck for more than 15–30 minutes, manually stop it and start it again.

2. Check Databricks status page
Confirm whether there is an active incident or degradation in your cloud/region.

3. Try a different warehouse size
Temporarily switch from Small to Medium and start it again. Sometimes a different size can provision successfully if a specific capacity pool is constrained.

4. Review Events / cluster logs
Check the warehouse events to see whether the issue is capacity, cloud provider quota, network, policy, or permissions related.

6. Open a Databricks support ticket
In practice, if a SQL Warehouse is stuck in Starting for more than 20–30 minutes, I would stop/restart first. If it repeats, try resizing or use another warehouse, then raise it with Databricks support.