Re: Insert speed new post

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Terry Fielder <terry(at)ashtonwoodshomes(dot)com>
Cc: Postgres General <pgsql-general(at)postgresql(dot)org>
Subject: Re: Insert speed new post
Date: 2007-07-02 20:55:10
Message-ID: 22615.1183409710@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Terry Fielder <terry(at)ashtonwoodshomes(dot)com> writes:
> Under pg 7.x the system performed fine.

> In 8.1.9, the insert statements seem to take a long time sometimes, upto
> several seconds or more.

> There is no primary key, but the table is never updated, only inserted.
> I removed the only index, with no improvement in performance (in case
> the 8.2 "resolves index locking issues" was the concern for an 8.1 install.

Hmm. With no indexes, inserts ought to be basically a constant-time
operation. I suspect what you are looking at is stalls caused by
checkpoints or other competing disk activity. I'd suggest watching the
output of "vmstat 1" or local equivalent, and seeing if you can
correlate the slow inserts with bursts of disk activity.

Have you tuned the 8.1 installation? I'm wondering about things like
checkpoint_segments and wal_buffers maybe being set lower than you had
them in 7.4.

regards, tom lane

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Poul Møller Hansen 2007-07-02 20:56:54 Invalid page header
Previous Message Tom Lane 2007-07-02 20:33:24 Re: recovery_target_time ignored or recovery always recovers to end of WAL