| From: | Thomas Lockhart <lockhart(at)alumni(dot)caltech(dot)edu> |
|---|---|
| To: | Steven Bradley <sbradley(at)llnl(dot)gov> |
| Cc: | pgsql-interfaces(at)postgreSQL(dot)org |
| Subject: | Re: [INTERFACES] Performance |
| Date: | 1999-06-24 02:02:23 |
| Message-ID: | 377191AF.BE5F54A@alumni.caltech.edu |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-interfaces |
> Is it unrealistic to expect Postgres to insert on the order of 100
> records per second on a Pentium 400 MHz/SCSI class machine running
> Linux? (Solaris on a comparable platform has about 1/2 the
> performance)
I'd expect roughly equivalent performance on different platforms. I
measured the insert performance a couple of years ago and found
roughly 50 inserts per second on my dual processor, 180MHz PPro
system, which was the same performance as I had on an Ingres
installation on a medium-speed Alpha/OSF-1 box.
Without knowing more about your memory and disk setup, I'll guess that
you can use more memory, can bump buffer size by using "-B 1024" or
something similar, and running without mandatory fsync() (use the "-F"
option). However, istm that your system should do better than 10ips
before you invoke "-F", otherwise you'll just be working around
another underlying problem/bottleneck.
Good luck.
- Thomas
--
Thomas Lockhart lockhart(at)alumni(dot)caltech(dot)edu
South Pasadena, California
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tim Allen | 1999-06-24 04:29:58 | Peculiar behaviour in libpq |
| Previous Message | Vadim Mikheev | 1999-06-24 01:21:15 | Re: [INTERFACES] Performance |