From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mescal.linbit (unknown [86.59.100.100]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by mail.linbit.com (LINBIT Mail Daemon) with ESMTP id CE4042DEAC78 for ; Wed, 7 Mar 2007 12:55:40 +0100 (CET) From: Philipp Reisner To: drbd-dev@lists.linbit.com Subject: Re: [Drbd-dev] Compression in DRBD Date: Wed, 7 Mar 2007 12:55:39 +0100 References: <1173262599.2536.25.camel@xenon> In-Reply-To: <1173262599.2536.25.camel@xenon> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-15" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Message-Id: <200703071255.40086.philipp.reisner@linbit.com> List-Id: Coordination of development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Am Mittwoch, 7. M=E4rz 2007 11:16 schrieb komal: > Hi all, > Bandwidth is an important issue in DRBD during replication of > data. If bandwidth is properly utilized then replication will be faster > and speed is what everyone needs. > When bandwidth utilization comes into picture, compressing data > is what > is needed. Earlier discussions were going on to use vtun with a tcp > tunnel with zlib compression factor 9 or so to achieve compression. > What about implementing compression in DRBD itself. Like using > zlib to > compress each packet and decompress it on receiver side in DRBD code > itself rather than using vtun and all. Wont it solve the purpose? > Hi, DRBD is a really complex piece of software, and I really do not want to duplicate anything that can get achieved on other levels. Compression on the network can for sure be achieved by other means=20 (outside of DRBD). =2Dphil =2D-=20 : Dipl-Ing Philipp Reisner Tel +43-1-8178292-50 : : LINBIT Information Technologies GmbH Fax +43-1-8178292-82 : : Vivenotgasse 48, 1120 Vienna, Austria http://www.linbit.com :