| Lists: | pgsql-performance |
|---|
| From: | "Merlin Moncure" <merlin(dot)moncure(at)rcsonline(dot)com> |
|---|---|
| To: | "Marc Cousin" <mcousin(at)sigma(dot)fr> |
| Cc: | <pgsql-performance(at)postgresql(dot)org>, "Qingqing Zhou" <zhouqq(at)cs(dot)toronto(dot)edu> |
| Subject: | Re: insert performance for win32 |
| Date: | 2005-11-02 13:36:57 |
| Message-ID: | 6EE64EF3AB31D5448D0007DD34EEB3417DD76C@Herge.rcsinc.local |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Lists: | pgsql-performance |
> I've done the tests with rc1. This is still as slow on windows ...
about
> 6-10
> times slower thant linux (via Ip socket). (depending on using prepared
> queries, etc...)
>
> By the way, we've tried to insert into the windows database from a
linux
> psql
> client, via the network. In this configuration, inserting is only
about 2
> times slower than inserting locally (the linux client had a slower CPU
> 1700Mhz agains 3000).
> Could it be related to a problem in the windows psql client ?
[OK, I'm bringing this back on-list, and bringing it to QingQing's
attention, who I secretly hope is the right person to be looking at this
problem :)]
Just to recap Marc and I have been looking at the performance disparity
between windows and linux for a single transaction statement by
statement insert on a very narrow table with no keys from a remote
client. Marc's observations showed (and I verified) that windows is
much slower in this case than it should be. I gprof'ed both the psql
client and the server during the insert and didn't see anything
seriously out of order...unfortunately QQ's latest win32 performance
tweaks haven't helped.
Marc's observation that by switching to a linux client drops time down
drastically is really intersing!
Merlin
| From: | Qingqing Zhou <zhouqq(at)cs(dot)toronto(dot)edu> |
|---|---|
| To: | Merlin Moncure <merlin(dot)moncure(at)rcsonline(dot)com> |
| Cc: | Marc Cousin <mcousin(at)sigma(dot)fr>, pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: insert performance for win32 |
| Date: | 2005-11-03 07:11:43 |
| Message-ID: | Pine.LNX.4.58.0511030153490.9091@eon.cs |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Lists: | pgsql-performance |
On Wed, 2 Nov 2005, Merlin Moncure wrote:
> >
> > By the way, we've tried to insert into the windows database from a
> > linux psql client, via the network. In this configuration, inserting
> > is only about 2 times slower than inserting locally (the linux client
> > had a slower CPU 1700Mhz agains 3000). Could it be related to a
> > problem in the windows psql client ?
> >
If you put client/server on the same machine, then we don't know how the
CPU is splitted. Can you take a look at the approximate number by
observing the task manager data while running?
If communication code is the suspect, can we measure the difference if we
disable the redefinition of recv()/send() etc in port/win32.h (may require
change related code a little bit as well). In this way, the socket will
not be able to pickup signals, but let see if there is any performance
difference first.
Regards,
Qingqing
>
> [OK, I'm bringing this back on-list, and bringing it to QingQing's
> attention, who I secretly hope is the right person to be looking at this
> problem :)]
>
P.s. You scared me ;-)