[Top] [All Lists]

Re: How many people regularly defrag their XFS filesystems?

To: Justin Piszcz <jpiszcz@xxxxxxxxxxxxxxx>
Subject: Re: How many people regularly defrag their XFS filesystems?
From: Bogdan Costescu <Bogdan.Costescu@xxxxxxxxxxxxxxxxxxxxx>
Date: Mon, 27 Mar 2006 12:43:35 +0200 (CEST)
Cc: KELEMEN Peter <Peter.Kelemen@xxxxxxx>, linux-xfs@xxxxxxxxxxx
In-reply-to: <Pine.LNX.4.64.0603260917001.17017@p34>
References: <Pine.LNX.4.64.0603251848140.15947@p34> <20060326114503.GA15588@xxxxxxxxxxxxxxxx> <Pine.LNX.4.64.0603260917001.17017@p34>
Sender: linux-xfs-bounce@xxxxxxxxxxx
On Sun, 26 Mar 2006, Justin Piszcz wrote:

Does anyone have any benchmarks for the XFS filesystem contrasted with the amount of fragmentation on the disk?

Not real benchmarks (which could be too specific), but a real world situation. I have at home a computer with a large PATA disk with XFS on it. This disk receives small amounts of data in time, from video encodes (own home productions and I like to experiment with filters a lot so data rate could be really low) or P2P applications (like recently from FC5 and CentOS 4.3 DVDs), either directly (run on the same computer) or via Samba.

Initially, this disk was formatted with ext3. However in time, I got to the point where the data rate when reading from disk was not enough to sustain a CD/DVD writing process. I switched it to XFS, but did not initially run xfs_fsr on a regular basis and this also started to behave the same after some time. After enabling xfs_fsr (nightly run), the reading speed is back to normal and the CD/DVD writing works fine. This is now in use for more than 2 years and I never looked back...

Bogdan Costescu

IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
E-mail: Bogdan.Costescu@xxxxxxxxxxxxxxxxxxxxx

<Prev in Thread] Current Thread [Next in Thread>