Received: from maia.hub.org (maia-3.hub.org [200.46.204.243]) by mail.postgresql.org (Postfix) with ESMTP id 02B93B5DC6C for ; Mon, 10 Oct 2011 13:14:53 -0300 (ADT) Received: from mail.postgresql.org ([200.46.204.86]) by maia.hub.org (mx1.hub.org [200.46.204.243]) (amavisd-maia, port 10024) with ESMTP id 99933-08 for ; Mon, 10 Oct 2011 16:14:46 +0000 (UTC) X-Greylist: from auto-whitelisted by SQLgrey-1.8.0-rc2 Received: from nm16.bullet.mail.ird.yahoo.com (nm16.bullet.mail.ird.yahoo.com [77.238.189.69]) by mail.postgresql.org (Postfix) with SMTP id 4A2CCB5DC33 for ; Mon, 10 Oct 2011 13:14:39 -0300 (ADT) Received: from [77.238.189.234] by nm16.bullet.mail.ird.yahoo.com with NNFMP; 10 Oct 2011 16:14:38 -0000 Received: from [212.82.108.241] by tm15.bullet.mail.ird.yahoo.com with NNFMP; 10 Oct 2011 16:14:38 -0000 Received: from [127.0.0.1] by omp1006.mail.ird.yahoo.com with NNFMP; 10 Oct 2011 16:14:38 -0000 X-Yahoo-Newman-Property: ymail-3 X-Yahoo-Newman-Id: 87.25422.bm@omp1006.mail.ird.yahoo.com Received: (qmail 57620 invoked by uid 60001); 10 Oct 2011 16:14:37 -0000 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yahoo.it; s=s1024; t=1318263277; bh=cf0cxLkFCvNTmRLRVVz8eZda5+yh0ASbAIb+v9q0tic=; h=X-YMail-OSG:Received:X-Mailer:References:Message-ID:Date:From:Reply-To:Subject:To:Cc:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=x5LNCYPkkS8SzQDaatR4poC2qUWriHYLKre5DA2dUDKGix5x7SPRYYD7aITUvJwoQ8Wk7ic8QKq6kr00TieTGlk7eDFI7sFd3qwYAFKvMGXTssJs4UgNChwSBRhm55g4mJgOCWsgJsmsfAyeNZajtz3Nlo/ScZUp+ls7CXygxNc= DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.it; h=X-YMail-OSG:Received:X-Mailer:References:Message-ID:Date:From:Reply-To:Subject:To:Cc:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=ftvZaw0770TU9eeee7ub6Fw/clX/XW+gW47NhQR/Ri5+P7+iRqxSOHa/CG1IpgkfMsGetI+KfE47pYfySLZ5cBDNIEdvur2f4vWNxC8JSQBfPGruUb0w9RoJ5xl+bdje/OYUV/HfBCSfH5GsYK/fF+2I+lAZ0kBJHgdOhaMD2ho=; X-YMail-OSG: 8z09A0cVM1lBo990g6YA0hbRjBNlNAXSdHu7lEIlM9JigHM adkbv.7s0ypGr23AjfKwRT2mXwIRPXVda9aO_RSGOiaU816vxKRKUS2FDz53 BwY5bo_Uaj4s3UtKfu6tWU8ly8nzAVK_VpZbpuD_oefRLlT5lpqns88iyVbb fDl4ATuDSdFWycFXEaAP5Y00dFsmoUHmQEtCTHBCOJ5S5agqBFRsStDcsHp2 zwjUktBIFe2rU0osGsnW_hGW0lsmPYmI2afGUOvrC.HBxHH0JkDyuE1Romdl d1LqgTh1P47Uz7sfW6lobY5CzbIEzV1txlt_wrTizVpToMFMhQdT6MZQ5gLz kfZhSxvJVMdGuCMY4ihLF1nNXIMK8SCtrxWsBAliHTm0ScaKhs5Zgmo2NGwG mN2Q- Received: from [85.33.62.89] by web29013.mail.ird.yahoo.com via HTTP; Mon, 10 Oct 2011 17:14:37 BST X-Mailer: YahooMailWebService/0.8.114.317681 References: <4E92F294.9010405@gmail.com> <4E92FB87.1030703@peak6.com> <1318259675.10344.YahooMailNeo@web29011.mail.ird.yahoo.com> <4E93119C.8090407@peak6.com> Message-ID: <1318263277.56967.YahooMailNeo@web29013.mail.ird.yahoo.com> Date: Mon, 10 Oct 2011 17:14:37 +0100 (BST) From: Leonardo Francalanci Reply-To: Leonardo Francalanci Subject: Re: Adding more memory = hugh cpu load To: "sthomas@peak6.com" Cc: alexandre - aldeia digital , "pgsql-performance@postgresql.org" In-Reply-To: <4E93119C.8090407@peak6.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable X-Virus-Scanned: Maia Mailguard 1.0.1 X-Spam-Status: No, hits=-1.899 tagged_above=-10 required=5 tests=BAYES_00=-1.9, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_NONE=-0.0001 X-Spam-Level: X-Archive-Number: 201110/83 X-Sequence-Number: 45305 > Then the=A0=0A=0A> database makes the fsync call, and suddenly the OS wan= ts to flush 2-6GB of data =0A> straight to disk. Without that background tr= ickle, you now have a flood that =0A> only the highest-end disk controller = or a backing-store full of SSDs or PCIe =0A> NVRAM could ever hope to absor= b.=0A=0A=0AIsn't=A0checkpoint_completion_target supposed to deal exactly wi= th that problem?=0A=0APlus: if 2-6GB is too much, why not decrease=A0checkp= oint_segments? Or=0Acheckpoint_timeout?=0A=0A> The kernel =0A> developers a= gree, or we wouldn't have dirty_bytes, or =0A> dirty_background_bytes, and = they wouldn't have changed the defaults to 5% =0A> and 10% instead of 10% a= nd 40%.=A0=0A=0A=0AI'm not saying that those kernel parameters are "useless= "; I'm saying they are used=0Ain =A0the same=A0way as the=A0checkpoint_segm= ents,=A0checkpoint_timeout and=0Acheckpoint_completion_target are used by p= ostgresql; and on a postgresql-only system=0AI would rather have postgresql= look after the fsync calls, not the OS.