Re: PostgreSQL pre-fork speedup

From: Tatsuo Ishii <t-ishii(at)sra(dot)co(dot)jp>
To: sdvmailer(at)yahoo(dot)com
Cc: pgman(at)candle(dot)pha(dot)pa(dot)us, pg(at)rbt(dot)ca, scott(dot)marlowe(at)ihs(dot)com, steve(at)blighty(dot)com, pgsql-hackers(at)postgresql(dot)org
Subject: Re: PostgreSQL pre-fork speedup
Date: 2004-05-13 13:00:41
Message-ID: 20040513.220041.74751033.t-ishii@sra.co.jp
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

> I ran the new Pgpool-1.2.2 and it was a bit faster on
> the TCP but still slower than on UNIX socket. I used
> the same script as before.
>
> TCP Socket (Pgpool 1.2.0)
> ----------
> 2.39 sec
>
> TCP Socket (Pgpool 1.2.2)
> ----------
> 0.80 sec
> 0.80 sec
> 0.79 sec
>
> UNIX Socket (Pgpool 1.2.2)
> -----------
> 0.026 sec
> 0.027 sec
> 0.027 sec
>
> Direct TCP connection (no pgpool)
> ---------------------------------
> 0.16 sec
> 0.15 sec
> 0.16 sec
>
>
> PgPool on TCP is still slower than direct connection
> but much faster than v1.2. Any other areas that can be
> improved?

This is strange. Using pgbench(pgbench -S -C -t 1000 -h localhost),
TCP socket with pgpool 1.2.2 runs about x2 faster than direct
connection.

Direct connection: 60TPS
With pgpool: 122TPS

Here is the set up:

Direct connection: pgbench <--TCP-->PG
With pgpool: pgbench <--TCP-->pgpool<--UNIX-->PG

Note: I use PostgreSQL 7.4.2. This means that pgpool forces pgbench to
fallback to V2 protocol (remember that pgpool does not support V3
yet), and the start up packet flys on the wire twice at the each
connection statge. This actually makes the benchmark worse, still
pgpool is better than direct connection however.
--
Tatsuo Ishii

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2004-05-13 13:15:21 Re: Probably security hole in postgresql-7.4.1
Previous Message Chris Bowlby 2004-05-13 12:11:19 dead index nodes...