Forum Discussion
Power BI refresh failed multiples times, and it got automatically fixed after few refreshes.
- 1 year ago
Hi SaiPrakash,
Thank you for reaching out to the Microsoft fabric community forum. Thank you bhanu_gautam, for your inputs on this issue.Based on the error message and code, the refresh failures are due to a transient internal service error in the Databricks Hive Metastore, which Power BI relies on when connecting to your dataset.
This typically happens when:
- The Databricks SQL warehouse is starting up (cold start)
- There is temporary unavailability or load on the Hive Metastore
- Power BI is trying to fetch Hive metadata (table info) and it times out or fails due to high resource demand
Since this resolved automatically after a few retries, it points to a transient condition. However, to reduce future occurrences:
- Make sure the SQL warehouse is pre-warmed or kept running before scheduled refreshes
- Prefer using direct SQL queries in Power BI instead of navigating through tables (which triggers Hive metadata calls)
- Stagger refresh schedules to reduce simultaneous load across datasets
- Ensure the Databricks ODBC driver is up to date (if a gateway is involved)
Kindly refer to the below link for better understanding:
Troubleshoot refresh scenarios - Power BI | Microsoft Learn
If this post helps, then please give us ‘Kudos’ and consider Accept it as a solution to help the other members find it more quickly.
Thank you for using Microsoft Community Forum.
SaiPrakash The error message you are encountering indicates that the refresh failure in Power BI is due to an internal service error related to the Databricks data source. Specifically, the error is caused by an issue with fetching Hive tables from the Hive database pfna_sem_glbl. The error class name org.apache.hadoop.hive.metastore.api.MetaException suggests that there is a problem with the Hive metastore, which is responsible for managing the metadata of Hive tables.
Ensure that the Databricks cluster is running smoothly and has sufficient resources (CPU, memory) to handle the queries.