Our Metabase instance is connected to a redshift database.
For some of the data we sync to redshift, we start by dropping the old table, re-creating it and then inserting the data all over again. It’s not the most efficient way, but it’s an easy way to capture deletes and schema changes.
My question is, what happens to the meta data that we have created in Metabase when the table it describes is dropped?
It would be sad to see all meta data go away every night
This is apparently not the case for postgres databases, at least for me, all meta data for a table is refetched when a table is dropped and created again with the same name.
This also causes the query builder for all saved cards to loose all information about the source table and fields. Raw sql cards are unaffected.