- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
06-05-2025 03:08 AM
Hey there,
today we cannot load/read data from Unity Catalog with the same cluster as we did yesterday successfully (no changes in clsuter configuration).
The error, which persists, according to the cluster logs is:
com.databricks.common.client.DatabricksServiceHttpClientException: TEMPORARILY_UNAVAILABLE: The service at /api/2.1/cluster-unity-catalog/catalogs/<catalog_name> is temporarily unavailable. Please try again later.
If is there any problem in Unity service, why is not appearing in https://status.databricks.com/?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
06-05-2025 04:17 AM
We are in the middle of migration from hive to unity....any news here?
The error still persists
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
06-05-2025 06:34 AM
Hello @GeoPer!
Could you please confirm which region your workspace is in? Also, is your cluster assigned to a specific group or access mode (shared or single user)?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
06-05-2025 06:47 AM
hey @Advika,
thank you for the reply.
The workspace is in eu-central-1 and the access mode of all-purpose cluster is Dedicated (formerly: Single user) to a group.
Yesterday with the same cluster - same configurations, everything was working like a charm.
Today we can only work with serverless or job clusters.
The problem still persists in All-purpose cluster, whith the above configurations.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
06-05-2025 07:05 AM
@Advika the issue is gone.
Now without any change all-purpose has access again to unity catalog.
Who knows what happened...
Thanks again for your interest
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
06-05-2025 07:19 AM
It seems this was a known issue with dedicated clusters. Thank you for confirming that it is working now!
