Received: (from majordomo@localhost)
	by oss.sgi.com (8.11.2/8.11.3) id f89NwrO00371
	for linux-xfs-outgoing; Sun, 9 Sep 2001 16:58:53 -0700
Received: from ente.berdmann.de (frnk-d5141a34.dsl.mediaWays.net [213.20.26.52])
	by oss.sgi.com (8.11.2/8.11.3) with SMTP id f89Nwmd00352
	for <linux-xfs@oss.sgi.com>; Sun, 9 Sep 2001 16:58:48 -0700
Received: from apollo.berdmann.de ([192.168.1.2] helo=berdmann.de)
	by ente.berdmann.de with esmtp (Exim 3.22 #1)
	id 15gETJ-00030Q-00; Mon, 10 Sep 2001 01:58:41 +0200
Message-ID: <3B9C0231.67E2BF01@berdmann.de>
Date: Mon, 10 Sep 2001 01:58:41 +0200
From: "Bernhard R. Erdmann" <be@berdmann.de>
X-Mailer: Mozilla 4.78 [en] (X11; U; Linux 2.4.10-pre4-xfs i586)
X-Accept-Language: de, en, fr
MIME-Version: 1.0
To: Dirk Wetter <dirkw@rentec.com>
CC: Linux XFS Mailing List <linux-xfs@oss.sgi.com>
Subject: Re: 10minutes for rm -rf on 400MB
References: <E15g8Dv-0006mt-00@ente.berdmann.de> <3B9BF400.1050900@rentec.com>
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
Sender: owner-linux-xfs@oss.sgi.com
Precedence: bulk

>we've been running XFS on the data disks of our HPC Linux cluster since <br>
>a while. we are quite happy with xfs, thx guys for your work!<br>
>the setup is:<br>
><br>
>- dual &gt;=1GHZ box, 4GB mem<br>
>- lvm 0.9beta7, phys. volume size ~140 GB, logical vol for xfs: 100GB<br>
>- no additional mount options or options for mkfs.xfs were given <br>
>- kernel 2.4.8pre4-xfs, highly patched SuSE 7.0 (not that it should matter)<br>
><br>
>a user complained that <i>rm -rf of 400MB&nbsp;</i> takes ~10 minutes (!) until
>the <br>
>command returns, whereas on the systems with reiserfs we have e.g. it <br>
>takes seconds.<br>

Some very important data is missing:
- what's the I/O performance of the disk subsystem?
- what was the system doing during the observed 10 min?

CPU power doesn't count as much as disk I/O performance because
unlink(2) on XFS is a synchronous operation.

>i don't know so much about the quality of the data, my guess is that some
>files <br>
>are small (~100k), others a big (a few hundred MB). i read in the FAQ that<br>
>XFS isn't particular good in rm-rf'ing&nbsp; files, which isn't really <b>the
></b>issue for<br>
>us, because in 99.9%of the time data is being read from the volume and not
><br>
>removed via rm -rf.<br>

So, three files à 100 MB and 1,024 files à 100 KB are 400 MB in sum and
even a busy system shouldn't take 10 min for deleting 1,027 files. I
guess your estimate of the file sizes is slightly wrong.