xfs
[Top] [All Lists]

Re: About removing files

To: linux-xfs <linux-xfs@xxxxxxxxxxx>
Subject: Re: About removing files
From: yoros@xxxxxxxxxx
Date: Fri, 25 Oct 2002 00:56:14 +0200
In-reply-to: <1035497418.31455.13.camel@jen.americas.sgi.com>
References: <20021024220639.GA20279@morpheus.matrix.com> <1035497418.31455.13.camel@jen.americas.sgi.com>
Sender: linux-xfs-bounce@xxxxxxxxxxx
User-agent: Mutt/1.4i
On Thu, Oct 24, 2002 at 05:10:18PM -0500, Steve Lord wrote:
> Depends which version of the code you are running, and how you
> create the iso. If your file is really fragmented then it might
> well take a while. If you are running the 1.1 release or earlier
> then you are probably running code which has some synchronous
> disk writes in the delete path. This is no longer the case in
> the latest code.
> 
> Try running xfs_bmap on your iso file, this outputs how the
> file is layed out on disk. If it is in lots of small chunks
> then that is part of the problem and you might want to look
> at how you create the iso images (and tell us how you do it too).

Good, I have applied xfs_bmap to a few files... I get a lot of extents
in some files and only a few of that in others.

xfs_bmap -l xxxx.avi | wc -l ----> 48800
xfs_bmap -l xxxx.iso | wc -l ---->   115

Now I don't have a most-fragmented ISO, but I found a most-fragmented
file. That file has about 20-40-50 blocks for each extent.

Here is my questions: ¿How can I obtain better performance? ¿Does the
number of blocks for extent related to creating/writting method?

Thank you for all this information, I am learning...

Pedro

-- 
Pedro Martinez Juliá
\  yoros@xxxxxxxx
)|    yoros@xxxxxxxxxx
/        http://yoros.cjb.net
Socio HispaLinux #311
Usuario Linux #275438 - http://counter.li.org
GnuPG public information:  pub  1024D/74F1D3AC
Key fingerprint = 8431 7B47 D2B4 5A46 5F8E  534F 588B E285 74F1 D3AC

Attachment: pgpem2BXdbZMw.pgp
Description: PGP signature

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