From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | "Marc Morin" <marc(at)sandvine(dot)com> |
Cc: | "Markus Schaber" <schabi(at)logix-tt(dot)com>, pgsql-performance(at)postgresql(dot)org |
Subject: | Re: Decreasing BLKSZ |
Date: | 2006-09-25 22:27:26 |
Message-ID: | 16132.1159223246@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
"Marc Morin" <marc(at)sandvine(dot)com> writes:
> No, an insert consists of roughly 10,000+ rows per transaction block.
Perhaps it would help to pre-sort these rows by key?
Like Markus, I'm pretty suspicious of lowering BLCKSZ ... you can try it
but it's likely to prove counterproductive (more btree index levels,
more rows requiring toasting, a tighter limit on what rows will fit at
all). I doubt I'd try to make it lower than a couple K in any case.
The bottom line here is likely to be "you need more RAM" :-(
I wonder whether there is a way to use table partitioning to make the
insert pattern more localized? We'd need to know a lot more about your
insertion patterns to guess how, though.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Mark Lewis | 2006-09-25 22:28:24 | Re: Decreasing BLKSZ |
Previous Message | Marc Morin | 2006-09-25 21:54:10 | Re: Decreasing BLKSZ |