I have a very basic scenario where I am using my custom catalog and want materialize view to get created from DLT table at the end of pipeline. The SQL used as below for the same.
where "loom_data_transform" is a Streaming table. But pipeline fails at the last stage where materialize view is tried to create from "loom_data_transform" with following error message.
org.apache.spark.sql.AnalysisException: '__materialization_mat_loom_data_report_1' was read as a stream (i.e. using `readStream` or `STREAM(...)`), but '__materialization_mat_loom_data_report_1' is not a streaming table. Either add the STREAMING keyword to the CREATE clause or read the input as a table rather than a stream.
Could someone please help me on this issue.
Finally, I able to resolve it but the fix is really surprising. When Materialize view need to be created from stream table as base, the syntax should be
live.loom_data_transform
in place of
STREAM(live.loom_data_transform)