Return-Path: Delivered-To: apmail-cassandra-user-archive@www.apache.org Received: (qmail 59851 invoked from network); 26 Apr 2010 19:12:01 -0000 Received: from unknown (HELO mail.apache.org) (140.211.11.3) by 140.211.11.9 with SMTP; 26 Apr 2010 19:12:01 -0000 Received: (qmail 68136 invoked by uid 500); 26 Apr 2010 19:12:00 -0000 Delivered-To: apmail-cassandra-user-archive@cassandra.apache.org Received: (qmail 68109 invoked by uid 500); 26 Apr 2010 19:12:00 -0000 Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: user@cassandra.apache.org Delivered-To: mailing list user@cassandra.apache.org Received: (qmail 68101 invoked by uid 99); 26 Apr 2010 19:12:00 -0000 Received: from athena.apache.org (HELO athena.apache.org) (140.211.11.136) by apache.org (qpsmtpd/0.29) with ESMTP; Mon, 26 Apr 2010 19:12:00 +0000 X-ASF-Spam-Status: No, hits=-0.2 required=10.0 tests=AWL,RCVD_IN_DNSWL_NONE,SPF_PASS X-Spam-Check-By: apache.org Received-SPF: pass (athena.apache.org: domain of ryan@twitter.com designates 209.85.221.192 as permitted sender) Received: from [209.85.221.192] (HELO mail-qy0-f192.google.com) (209.85.221.192) by apache.org (qpsmtpd/0.29) with ESMTP; Mon, 26 Apr 2010 19:11:55 +0000 Received: by qyk30 with SMTP id 30so15226513qyk.16 for ; Mon, 26 Apr 2010 12:11:34 -0700 (PDT) MIME-Version: 1.0 Received: by 10.229.191.76 with SMTP id dl12mr5325222qcb.97.1272309093855; Mon, 26 Apr 2010 12:11:33 -0700 (PDT) Received: by 10.229.211.78 with HTTP; Mon, 26 Apr 2010 12:11:33 -0700 (PDT) In-Reply-To: References: Date: Mon, 26 Apr 2010 12:11:33 -0700 Message-ID: Subject: Re: Can Cassandra make real use of several DataFileDirectories? From: Ryan King To: user@cassandra.apache.org Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable 2010/4/26 Roland H=E4nel : > Hm... I understand that RAID0 would help to create a bigger pool for > compactions. However, it might impact read performance: if I have several > CF's (with their SSTables), random read requests for the CF files that ar= e > on separate disks will behave nicely - however if it's RAID0 then a rando= m > read on any file will create a random read on all of the hard disks. > Correct? Without RAID0 you will end up with host spots (a compaction could end up putting a large SSTable on one disk, while the others have smaller SSTables). If you have many CFs this might average out, but it might not and there are no guarantees here. I'd reccomend RAID0 unless you have reason to do something else. -ryan