Good morning, guys.
I’m trying to create a reflection that it’s pretty big. This VDS was made up from a query which joins other 6 VDS (~12 millions registers each). For each of these 6 VDS I’ve created a reflection as well. However, when I try to create a reflection for the main VDS (which joins the other 6), I got a memory error. When I check the profile, seems that Dremio consumed a lot of memory in operators like HASH_JOIN (which I know that consumes a lot of memory). My question is: How can I handle problems like this, when I need to create this kind of reflection?
Here’s my profile: c7ec2295-879d-4a40-a1b6-bb04fb6f9db7.zip (397.7 KB)
I’m running Dremio in a cluster with 1 coordinator and 2 executors: