Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Once a materialized lake view (MLV) is created, the service can handle its future refreshes based on the schedule provided in the generated lineage.
The following refresh operations can occur depending on the updates to the data in the source tables.
Full Refresh: A full refresh entails evaluating the complete dataset of the source tables whenever any modifications are detected in the source tables.
No Refresh: If the source tables remain unchanged, the MLV refresh is skipped, which saves unnecessary processing and reduces costs.
Refresh a materialized lake view using Spark SQL command
If it's necessary to quickly reflect changes in an MLV, you can utilize the following command to refresh the MLV.
REFRESH MATERIALIZED LAKE VIEW [workspace.lakehouse.schema].MLV_Identifier [FULL]
Argument:
FULL: It's an optional argument. If the FULL keyword is used, a full refresh of the MLV is performed. If omitted, the system decides whether to run a full refresh or skip it based on the source data.
Note
Refreshing an MLV that uses non-delta tables as its source initiate a full refresh of the MLV.
Known issues
- At present, all refresh operations default to a full refresh.