Virtual datasets


I always create virtual datasets with reflections for my queries in different databases. Is this good practice? In addition, I am facing problems with queries, when I run with the Dremio query run, it works and very fast (that is, a few minutes). However, after a few hours of running again, the query takes a long time (that is, many hours). What can I do to improve this?



Can you please provide us with the profiles of the slow and fast runs?

Share a query profile