CPU thrashing issue?

Hi All,

We are trying to use metabase to do some data science stuff on a postgres database. We have a docker image installed on AWS giving the java heap 24G of memory while leaving everything else as default aside from making the metastore point to postgres instead of H2. What we are observing is that when we trigger a sync for the database, during the process metabase will eventually hang and error out with a 504 error on the application. We don’t seem to be getting any information from the logs since they seem to be silent during this time. Is there anyway to make the logs more verbose? In the aws instance we are seeing that metabase is pegging one cpu at 100%, is there a way to get the application to split resource usage amongst more than one cpu?

Another question is that the initial setup of the metabase and the postgres instance, it looks like there is a partial sync that occurred because we can definitely see the tables and choose them when creating a question (though not all the tables appear), but we cannot see them in the data model tab in the admin panel (we want to basically hide some of these tables from appearing in the question).

Note: Our postgres instance has over 11K tables (they are not really too big, just a lot of tables).

Similar issues I’m having but on CSV download of wide table