From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mx1.redhat.com (ext-mx04.extmail.prod.ext.phx2.redhat.com [10.5.110.8]) by int-mx05.intmail.prod.int.phx2.redhat.com (8.13.8/8.13.8) with ESMTP id n9RGXK55026552 for ; Tue, 27 Oct 2009 12:33:20 -0400 Received: from mail-iw0-f194.google.com (mail-iw0-f194.google.com [209.85.223.194]) by mx1.redhat.com (8.13.8/8.13.8) with ESMTP id n9RGWxBv013959 for ; Tue, 27 Oct 2009 12:33:02 -0400 Received: by iwn32 with SMTP id 32so248375iwn.23 for ; Tue, 27 Oct 2009 09:32:59 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: <4AE711B2.1060800@worldspice.net> References: <4AE711B2.1060800@worldspice.net> Date: Tue, 27 Oct 2009 08:32:59 -0800 Message-ID: <1c748a490910270932i4358d850va6d3a232775950dc@mail.gmail.com> Subject: Re: [linux-lvm] Quick Question RE Snapshots From: Larry Dickson Content-Type: multipart/alternative; boundary=001636920ca93605f80476ed3b84 Reply-To: LVM general discussion and development List-Id: LVM general discussion and development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , List-Id: To: LVM general discussion and development --001636920ca93605f80476ed3b84 Content-Type: text/plain; charset=ISO-8859-1 In my experience yes - if you are doing long sequential writes, snapshot turns them into random IO. If you were doing short random writes to begin with, then snapshot should not have such a huge relative effect, because the non-snapshot performance is already slow. Larry Dickson Cutting Edge Networked Storage On 10/27/09, Ryan Anderson wrote: > I'm sure this has been covered before, but I didn't see it in the > archive: Is a near 90% drop in write performance "normal" snapshot > behavior? If so, :( > > If not, are there any good resources I should be looking at? Every > search I perform comes up with one of two types of articles on the matter: > > > 1) Snapshots are awesome! They let you...(list of really nifty things) > > 2) Snapshots suck! My system is completely useless while they're active! > > -- > Ryan Anderson > (901) 843 9300 > Systems Engineer > WorldSpice Technologies > > > _______________________________________________ > linux-lvm mailing list > linux-lvm@redhat.com > https://www.redhat.com/mailman/listinfo/linux-lvm > read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/ > > --001636920ca93605f80476ed3b84 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable
In my experience yes - if you are doing long sequential writes, snapsh= ot turns them into random IO. If you were doing short random writes to begi= n with, then snapshot should not have such a huge relative effect, because = the non-snapshot performance is already slow.
=A0
Larry Dickson
Cutting Edge Networked Storage

On 1= 0/27/09, Ryan Anderson <ryan@worldspice.net> wrote:
I'm sure this has been cover= ed before, but I didn't see it in the
archive: Is a near 90% drop in= write performance "normal" snapshot
behavior? If so, :(

If not, are there any good resources I should be= looking at? Every
search I perform comes up with one of two types of ar= ticles on the matter:


1) Snapshots are=A0=A0awesome! They let yo= u...(list of really nifty things)

2) Snapshots suck! My system is completely useless while they're ac= tive!

--
Ryan Anderson
(901) 843 9300
Systems Engineer
W= orldSpice Technologies


_________________________________________= ______
linux-lvm mailing list
linux-lvm= @redhat.com
https://www.redhat.com/mailman/listinfo/linux-lvm
read the LVM= HOW-TO at http://tldp.org/HOW= TO/LVM-HOWTO/


--001636920ca93605f80476ed3b84--