X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.2]) by svr1.postgresql.org (Postfix) with ESMTP id F03A3D1D25B for ; Fri, 23 Apr 2004 19:04:28 -0300 (ADT) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (neptune.hub.org [200.46.204.2]) (amavisd-new, port 10024) with ESMTP id 93235-09 for ; Fri, 23 Apr 2004 19:04:27 -0300 (ADT) Received: from mta4.rcsntx.swbell.net (mta4.rcsntx.swbell.net [151.164.30.28]) by svr1.postgresql.org (Postfix) with ESMTP id 7E113D1D08F for ; Fri, 23 Apr 2004 19:04:24 -0300 (ADT) Received: from one.cheapcomplexdevices.com (adsl-68-122-9-222.dsl.pltn13.pacbell.net [68.122.9.222]) by mta4.rcsntx.swbell.net (8.12.10/8.12.10) with ESMTP id i3NM3xY4004066; Fri, 23 Apr 2004 17:04:00 -0500 (CDT) Received: from localhost (one [127.0.0.1]) by one.cheapcomplexdevices.com (Postfix) with ESMTP id D39AA216F7D; Fri, 23 Apr 2004 15:10:20 -0700 (PDT) Date: Fri, 23 Apr 2004 15:10:20 -0700 (PDT) From: Ron Mayer X-X-Sender: mayer@one To: Manfred Koizar Cc: Pallav Kalva , pgsql-performance@postgresql.org Subject: Re: Setting Shared Buffers , Effective Cache, Sort Mem In-Reply-To: Message-ID: References: <4088062E.2050901@deg.cc> <4089261A.9070700@deg.cc> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Virus-Scanned: by amavisd-new at postgresql.org X-Spam-Status: No, hits=0.1 tagged_above=0.0 required=5.0 tests=RCVD_IN_SORBS X-Spam-Level: X-Archive-Number: 200404/419 X-Sequence-Number: 6719 On Fri, 23 Apr 2004, Manfred Koizar wrote: > > Setting shared_buffers to half your available memory is the worst thing > you can do. You would end up caching exactly the same set of blocks in > the internal buffers and in the OS cache, thus effectively making one of > the caches useless. One minor detail... You wouldn't really cache the _exact_ same blocks because cache-hits in shared-buffers (on the most frequently accessed pages) would let the OS cache some other pages in it's cache. But in my experience Manfred's right that there's no benefit and some penalty to making shared_buffers so large it takes a significant piece away from the OS's caching.