X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id D21CD32A1FF for ; Wed, 20 Oct 2004 14:36:57 +0100 (BST) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024) with ESMTP id 17237-09 for ; Wed, 20 Oct 2004 13:36:46 +0000 (GMT) Received: from mproxy.gmail.com (rproxy.gmail.com [64.233.170.194]) by svr1.postgresql.org (Postfix) with ESMTP id B2670329F90 for ; Wed, 20 Oct 2004 14:36:48 +0100 (BST) Received: by mproxy.gmail.com with SMTP id 74so454090rnk for ; Wed, 20 Oct 2004 06:36:49 -0700 (PDT) DomainKey-Signature: a=rsa-sha1; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:reply-to:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:references; b=W37flJuiyWw8N8vsFdmSEhpLS1CCwtm6PzuC7TKPyA4aDHp6nitkJdRIYbgxDR5mGzM+L8a9S7MISuNB2bgPMMchEbB50ikYB1uhhRDfEM0VzkFf3ng6VBLtqzOznHQoSp5pwDMal7aowkYCdZz2y0P5YrVO1BiQky7QgfgO5qI Received: by 10.38.152.39 with SMTP id z39mr2542808rnd; Wed, 20 Oct 2004 06:36:49 -0700 (PDT) Received: by 10.38.76.42 with HTTP; Wed, 20 Oct 2004 06:36:49 -0700 (PDT) Message-ID: <4a0cafe204102006361e372071@mail.gmail.com> Date: Wed, 20 Oct 2004 08:36:49 -0500 From: Josh Close Reply-To: Josh Close To: POSTGRES-PERFORMANCE Subject: Re: how much mem to give postgres? In-Reply-To: <29976.1098246931@sss.pgh.pa.us> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit References: <4a0cafe2041019153133f854e9@mail.gmail.com> <29976.1098246931@sss.pgh.pa.us> X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.0 tagged_above=0.0 required=5.0 tests= X-Spam-Level: X-Archive-Number: 200410/315 X-Sequence-Number: 8791 On Wed, 20 Oct 2004 00:35:31 -0400, Tom Lane wrote: > I suspect that fooling with shared_buffers is entirely the wrong tree > for you to be barking up. My suggestion is to be looking at individual > queries that are slow, and seeing how to speed those up. This might > involve adding indexes, or tweaking the query source, or adjusting > planner parameters, or several other things. EXPLAIN ANALYZE is your > friend ... > > regards, tom lane Only problem is, a "select count(1)" is taking a long time. Indexes shouldn't matter with this since it's counting every row, right? The tables are fairly well indexed also, I could probably add a few more. If shared_buffers isn't the way to go ( you said 10k is the sweetspot ), then what about the effective_cache_size? I was suggested on the general list about possibly setting that to 75% of ram. Thanks. -Josh