<div dir="ltr"><div class="gmail_quote"><div>Hi Stefan,<br><br>We have been trying to optimize Glusterfs on our network and our results match your previous results. After reading your post, I tried removing AFR but our performances are still poor as compared to NFS (particulalrly for small files). Could you please post your new volume spec file and/or any other improvements that you made to any of the layers? That'll be very helpful.<br>
<br>Thanks and regards<br>Chandranshu<br> </div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"> Date: Thu, 18 Sep 2008 10:47:22 +0200<br>
From: <a href="mailto:stefan@mdy.univie.ac.at">stefan@mdy.univie.ac.at</a> (Stefan Boresch)<br>
Subject: Re: [Gluster-users] Poor performance with AFR<br>
To: <a href="mailto:gluster-users@gluster.org">gluster-users@gluster.org</a><br>
Message-ID: <<a href="mailto:20080918084722.GH9860@loop.mdy.univie.ac.at">20080918084722.GH9860@loop.mdy.univie.ac.at</a>><br>
Content-Type: text/plain; charset=us-ascii<br>
<br>
Dear everyone,<br>
<br>
thank you for your replies. I have some additional data that have<br>
clarified issues for me a bit:<br>
<br>
I have repeated my tests without AFR (basically replicating the plain<br>
NFS setup)<br>
<br>
<br>
> > CP-A MAKE<br>
> > local disk < 5sec < 0.3sec<br>
> > NFS (100MBit) 55sec+-2sec < 2sec<br>
> > glusterfs (I) 4m29sec 17sec<br>
> > glusterfs (II) 4m05sec 18sec<br>
glusterfs w/o AFR 45+-2 sec 9sec <==NEW<br>
<br>
So, most of the poor performance is due to AFR. Note that the<br>
copy actually is now faster than NFS. Interestingly, make<br>
still runs much slower (although compared to the actual<br>
compile time, this overhead should be negligible in practice)<br>
<br>
Also, upon running some NFS benchmarks between the two servers, I<br>
noted some strange results, letting me suspect some creeping hardware<br>
issues.<br>
<br>
So, I guess I'll (a) wait for glusterfs 1.4.x and (b) look out for<br>
some better hardware to test things in the meantime.<br>
<br>
Sorry if I caused confusion; I should have checked some of these things<br>
earlier.<br>
<br>
Best regards,<br>
<br>
Stefan Boresch<br>
<br>
--<br>
Stefan Boresch<br>
Institute for Computational Biological Chemistry<br>
University of Vienna, Waehringerstr. 17 A-1090 Vienna, Austria<br>
Phone: -43-1-427752715 Fax: -43-1-427752790<br>
<br></blockquote></div></div>