Re: Random plan selection in DBT-3

From: ITAGAKI Takahiro <itagaki(dot)takahiro(at)oss(dot)ntt(dot)co(dot)jp>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: Random plan selection in DBT-3
Date: 2008-01-07 03:06:49
Message-ID: 20080107120132.6486.ITAGAKI.TAKAHIRO@oss.ntt.co.jp
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers


Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:

> ITAGAKI Takahiro <itagaki(dot)takahiro(at)oss(dot)ntt(dot)co(dot)jp> writes:
> > I tested 8.3beta4 with DBT-3 (TPC-H) and found unstable selection of plans.
> > Planner randomly selected two types of plans (A, B) when I repeated EXPLAIN.
>
> Are you sure the statistics weren't changing underneath you (due to
> auto-vacuum or auto-analyze)?

Sure. There were no statistic changes.

There might be regularity in the plan selection. When I connect to postgres
newly, I see plans { A, B, A, A ... }. Then disconnect, re-connect and send
the same EXPLAINs, the same pattern was reproduced.

Regards,
---
ITAGAKI Takahiro
NTT Open Source Software Center

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2008-01-07 03:26:42 Re: Random plan selection in DBT-3
Previous Message Tom Lane 2008-01-07 02:54:49 Re: Random plan selection in DBT-3