xfs
[Top] [All Lists]

Re: benchmarks

To: "P.Dixon" <P.Dixon@xxxxxxxxx>
Subject: Re: benchmarks
From: Utz Lehmann <leh@xxxxxxxxxxxxxx>
Date: Fri, 13 Jul 2001 14:19:22 +0200
Cc: linux-xfs@xxxxxxxxxxx
In-reply-to: <Pine.LNX.4.33.0107130921430.29969-100000@xxxxxxxxxxxxxxxxxxx>; from P.Dixon@xxxxxxxxx on Fri, Jul 13, 2001 at 09:23:15AM +0100
References: <Pine.LNX.4.33.0107130921430.29969-100000@xxxxxxxxxxxxxxxxxxx>
Sender: owner-linux-xfs@xxxxxxxxxxx
User-agent: Mutt/1.2.5i
Hi

My experience with fs benchmarking is, that under different conditions the
results are different (and another filesystem wins).

Nearly almost fs benchmark have nothing todo with real life. They are run on
a freshly formated partition and so on...

For the reiser vs. xfs case just make following:

- Format a partition
- Fillup the partition with garbage, lots of differnet sized files and
  simultaneous write (just a few untars in parallel)
- Delete some files
- Repeat the last 2 steps a few times
- After this, the filesystem should filled 97-99% including the comming 
benchmark
- Run the benchmark

My guess: xfs will outperform reiser (and every other linux fs)

Then publish the results on a webpage with the headline "xfs is 10 times
faster as reiserfs!"


utz


P.Dixon [P.Dixon@xxxxxxxxx] wrote:
> Hello,
> 
> Any idea why xfs appears to be very much slower than reiserfs with these
> benchmarks:
> 
> http://www.namesys.com/benchmarks/mongo/2.4.5-xfs-ext2_vs_reiserfs.html
> 
> Admittedly, the benchmarks were done by namesys...
> 
> -- 
> 
>                                       Paul
> 
> -------------------------------------------------------------------------------
> Paul Dixon                           Email: P.Dixon@xxxxxxxxx
> Department of Physics                Phone: (020) 7882 5054
> Queen Mary, University of London     Fax  : (020) 7882 5054
> Mile End Road, London E1 4NS         URL  : http://hepwww.ph.qmw.ac.uk/~pd
> -------------------------------------------------------------------------------


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