Return-Path: Delivered-To: apmail-cassandra-user-archive@www.apache.org Received: (qmail 8993 invoked from network); 11 May 2010 18:40:04 -0000 Received: from unknown (HELO mail.apache.org) (140.211.11.3) by 140.211.11.9 with SMTP; 11 May 2010 18:40:04 -0000 Received: (qmail 95901 invoked by uid 500); 11 May 2010 18:40:03 -0000 Delivered-To: apmail-cassandra-user-archive@cassandra.apache.org Received: (qmail 95884 invoked by uid 500); 11 May 2010 18:40:03 -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 95876 invoked by uid 99); 11 May 2010 18:40:03 -0000 Received: from nike.apache.org (HELO nike.apache.org) (192.87.106.230) by apache.org (qpsmtpd/0.29) with ESMTP; Tue, 11 May 2010 18:40:03 +0000 X-ASF-Spam-Status: No, hits=3.6 required=10.0 tests=FS_REPLICA,SPF_PASS X-Spam-Check-By: apache.org Received-SPF: pass (nike.apache.org: domain of scode@spotify.com designates 209.85.161.44 as permitted sender) Received: from [209.85.161.44] (HELO mail-fx0-f44.google.com) (209.85.161.44) by apache.org (qpsmtpd/0.29) with ESMTP; Tue, 11 May 2010 18:39:55 +0000 Received: by fxm7 with SMTP id 7so718483fxm.31 for ; Tue, 11 May 2010 11:39:35 -0700 (PDT) MIME-Version: 1.0 Received: by 10.239.162.136 with SMTP id l8mr594949hbd.212.1273603175250; Tue, 11 May 2010 11:39:35 -0700 (PDT) Received: by 10.239.152.4 with HTTP; Tue, 11 May 2010 11:39:35 -0700 (PDT) In-Reply-To: References: <4BE9815D.1060208@dehora.net> Date: Tue, 11 May 2010 20:39:35 +0200 Message-ID: Subject: Re: replication impact on write throughput From: =?UTF-8?Q?Peter_Sch=C3=BCller?= To: user@cassandra.apache.org Content-Type: text/plain; charset=UTF-8 X-Virus-Checked: Checked by ClamAV on apache.org > The biggest impact on your write performance will most likely be the > consistency level of your writes. In other words, how many nodes you want to > wait for before you acknowledge the write back to the client. I believe the consistency level is only expected to have a significant impact on latency (and throughput if you have a client who's write throughput is limited by latency). Since the writes will still happen at the appropriate replication level, the total number of writes having to be processed by the cluster should remain the same. -- / Peter Schuller aka scode