Redshift - "ERROR: type "e" does not exist" is back

Hi Metabase team,

I’ve been a long time sufferer of this GitHub issue, which had gone away for the last couple of months. Unfortunately however, it came back roughly around the time I upgraded to 0.33.5. I don’t know for sure that 0.33.5 had anything to do with the error coming back or not, but the timing is suspicious.

My main question here is what should I do next to help get this resolved? The aforementioned ticket is already closed. Should I comment on that ticket, make a new ticket, etc…?

A couple of other notes: I’m on Elastic Beanstalk. Restarting my app servers seems to temporarily fix the issue for a short time, but then it comes back. I mainly see the error when Metabase is trying to sync. Here’s a stack trace snippet:

[5041e072-3c43-41e6-b682-aff4639922af] 2019-11-19T11:00:00-08:00 INFO metabase.sync.util STARTING: step 'sync-tables' for redshift Database 2 'Panoply'
[5041e072-3c43-41e6-b682-aff4639922af] 2019-11-19T11:00:00-08:00 ERROR metabase.sync.util Error running sync step: ERROR: type "e" does not exist
"--> driver.sql_jdbc.sync$get_tables.invokeStatic(sync.clj:90)"

Hi @JZee
Since you’re also being hit by, then it might have something to do with that, but please make sure you’re running and not 0.33.5, since there were a build problem with that version, which means some people might be using a broken version of 0.33.4

OK, thanks for the heads up! I’ll sit tight until the smoke clears on before submitting a ticket. I’m on now.

1 Like

@JZee Just for your info, 0.33.6 is being released right now, but doesn’t contain any changes related to Redshift.