Postgres Pro
Downloads
Home   >   mailing lists

Re: Query plan prefers hash join when nested loop is much faster - Mailing list pgsql-general

From Michael Lewis
Subject Re: Query plan prefers hash join when nested loop is much faster
Date
Msg-id CAHOFxGpb2_KB34MZdBGeH52nrkqPUkvf78j+V_b9Vt1by64Tfg@mail.gmail.com
Whole thread Raw
In response to Query plan prefers hash join when nested loop is much faster  (iulian dragos <iulian.dragos@databricks.com>)
Responses Re: Query plan prefers hash join when nested loop is much faster  (iulian dragos <iulian.dragos@databricks.com>)
List pgsql-general
Tree view
Your system is preferring sequential scan to using test_result_module_result_id_idx in this case. What type of storage do you use, what type of cache hits do you expect, and what do you have random_page_cost set to? That comes to mind as a significant factor in choosing index scans based on costs.

pgsql-general by date:

Previous
From: iulian dragos
Date:
Subject: Query plan prefers hash join when nested loop is much faster
Next
From: greigwise
Date:
Subject: pgbouncer bug?