| From: | Qingqing Zhou <zhouqq(at)cs(dot)toronto(dot)edu> |
|---|---|
| To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | Merlin Moncure <merlin(dot)moncure(at)rcsonline(dot)com>, pgsql-performance(at)postgresql(dot)org, Marc Cousin <mcousin(at)sigma(dot)fr> |
| Subject: | Re: insert performance for win32 |
| Date: | 2005-11-04 07:29:49 |
| Message-ID: | Pine.LNX.4.58.0511040215540.22190@eon.cs |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
On Thu, 3 Nov 2005, Tom Lane wrote:
>
> On Unix I get a dead flat line (within measurement noise), both local
> loopback and across my LAN.
>
> after 50000 30.20 sec
> after 100000 31.67 sec
> after 150000 30.98 sec
> after 200000 29.64 sec
> after 250000 29.83 sec
>
Confirmed in Linux. And on a winxp machine(sp2) with server, client
together, with (see almost no performance difference) or without my "fake"
write, the observation is still hold for both cases:
after 50000 25.21 sec
after 100000 26.26 sec
after 150000 25.23 sec
after 200000 26.25 sec
after 250000 26.58 sec
In both cases, postgres 67% cpu, psql 15~20%, rest: system process. Kernel
time is 40+% -- where from?
Regards,
Qingqing
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Michal Taborsky | 2005-11-04 11:38:30 | Searching union views not using indices |
| Previous Message | Tom Lane | 2005-11-03 23:56:59 | Re: insert performance for win32 |