Meltano

Question about `target-postgres`
We've got a meltano container running that pulls from Airtable every 5 minutes and is landing the data in an Aurora Postgres DB.

In the window that the job runs, the DB is being hammered pretty hard. Looking at RDs metrics it seems that `autovacuum` is running constantly.

Has anyone encountered a similar problem, and have any advice, or suggests for some better, non-default config? Basically, I think we should not autovacuum quite so often...

I think the problem is related to the DBT transform that runs afterwards that is constantly  modifying schema, and this is why the PG catalogue tables are the ones being vacuumed

DB at 60% cpu seems fine to me, why is it an issue for you?

60% is fine if it's meaningful work.
0.5% worth of work and 59.5% just doing autovacuuming is a bit nuts IMO.

It's also costing extra in IO charges with AWS, which is unnecessary.

I think I can switch the `materialize` config for my DBT transforms from `table` to `view` which will cut down on DDL.

Does anyone know if Meltano itself triggers lots of DDL on it's raw tables? I guess that is target dependent...