From: | "David Wilson" <david(dot)t(dot)wilson(at)gmail(dot)com> |
---|---|
To: | "Alex Vinogradovs" <AVinogradovs(at)clearpathnet(dot)com> |
Cc: | "Joshua Drake" <jd(at)commandprompt(dot)com>, pgsql-general(at)postgresql(dot)org |
Subject: | Re: non-WAL btree? |
Date: | 2008-08-01 20:16:19 |
Message-ID: | e7f9235d0808011316r5fa30d98vfd2933e96af51c9a@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On Fri, Aug 1, 2008 at 4:07 PM, Alex Vinogradovs
<AVinogradovs(at)clearpathnet(dot)com> wrote:
> By loading in bulks, I mean I load some 40-50 thousand
> rows at once into a table that already has some millions.
> Index rebuild on that table after each 50k inserts will
> be even less efficient ;)
How many indexes do you have on this...? I do this pretty regularly
(actually, I do 4k batches with COPY, 4-10 concurrent batches every 10
seconds, for 2-3 days at a time) and, having testing dropping indices,
nothing to do with the index has a particularly strong performance
impact.
That said, a significant increase in checkpoint segments was required
to get good performance out of the above use case. If you haven't
tried that, I'd say that's a good place to start.
What makes you think it's specifically index WAL work, though?
--
- David T. Wilson
david(dot)t(dot)wilson(at)gmail(dot)com
From | Date | Subject | |
---|---|---|---|
Next Message | Joshua Drake | 2008-08-01 20:16:26 | Re: non-WAL btree? |
Previous Message | Alex Vinogradovs | 2008-08-01 20:07:18 | Re: non-WAL btree? |