I've increased the max results and max aggregated results to 1,000,000
Ordinarily when I run a particular stored proc with a comma-delimited string as input with 1,000+ entries it completes instantaneously when run via SQL Studio/Azure Data Studio etc. It returns 60k records and even when running 10,000 entires it completes in 30 seconds (600k records)...
When using Metabase using the exact same 1000+ inputs to the stored proc it takes minutes or may even timeout.
My IDE pulls all 600,000 in the last run I did within 5-10 seconds. I'm using Azure Data Studio and SQL Management Studio; these both pull in everything (and does it very fast).
I've got several terabytes of RAM on this server so I'm fine with devoting large quantities of RAM to Metabase.
edit: Also for reference - the same stored procedure pulling down 4.5 million records comes back in about 1 minute using Azure Data Studio or SQL Management Studio. Metabase won't even return it.
I'm running Metabase via docker and Postgres app database.
One other interesting point. I've seen queries that Metabase uses when querying via the database browser and they are very inefficient. Instead of doing an IN() when filtering, Metabase will do large numbers of OR operations which is massively inefficient.
I really want to make metabase work for this use-case.... it just doesn't seem ( so far) like it can handle what I consider to be large amounts of data
edit 2: I just ran the same 600k record query and it took Metabase 10 minutes to return results.