Is there a workaround for UUID's in postgres on secondary join

1

We are using a Postgres database as a datasource. We use uuids as pks in almost all of our tables.
Uuids where causing issues in the older versions, which could workaround using the ‘legacy dialect’ option.

‘ERROR: collations are not supported by type uuid’

However, now this option is gone, but the error is still there when using queries that join using an uuid.
select * table1 t1
inner join table2 t2 on t1.id = t2.id
inner join table3 t3 on t2.id = t3.id

The error happens with the second join, I’m guessing that Dremio doesn’t figure out this t2.id is a uuid and still adds the ordering with collate to the query.

Is there a workaround for this issue?

2

@markmeeus

Would like to see what the push down is, can you please attach the job profile?

Thanks
Bali