We are facing multiple issues trying to set up pre-aggregations. Our store is Big Query. We have tried both external pre-aggregations on Postgres as well as Cube Store, to no avail. For now, we are just trying to use a simple “originalSql” (ie copy the whole table).
Here are a sample of issues we get:
- when a cube dimension is defined with a custom SQL statement (ie in this case, BigQuery SQL), it seems that Cube tries to employ the same syntax in the pre-aggregation store, which fails because of incompatible SQL dialects.
- if a BigQuery table contains repeated fields, pre-aggregation fails with various errors
- various syntax errors related to extra brackets or parentheses, which are a little hard to debug because of terse error messages.
- “Invalid identifier xxx for schema” (only with Cube Store)
- "Invalid function xxx (with Cube Store)
We’d love to use pre-aggregations and we’d appreciate some help debugging these errors