<p dir="ltr">Ben,</p>
<p dir="ltr">I suspect meta-data / 'ls -l' performance is very important for my svn use-case.</p>
<p dir="ltr">Having said that, what do you mean by small file performance? I thought what people meant by this was really the overhead of meta-data, with a 'ls -l' being a sort of extreme case (pure meta-data).<br>
Obviously if you also have to read and write actual data (albeit not much at all per file), then the effect of meta-data overhead would get diluted to a degree, bit potentially still very present.</p>
<p dir="ltr">Would there be an easy way to tell how much time is spent on meta-data vs. Data in a profile output?</p>
<p dir="ltr">One thing I wonder: do your comments apply to both native Fuse and NFS mounts?</p>
<p dir="ltr">Finally, all this brings me back to my initial question really: are there any tuning recommendation of configuration tuning for my requirement (small file read/writes on a pair of nodes with replication) beyond the thread counts and lookup optimize?<br>
Or are those by far the most important in this scenario?<br></p>
<p dir="ltr">Thx,<br>
Thibault.</p>
<div class="gmail_quot<blockquote class=" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">----- Original Message -----<br>
> From: <a href="mailto:hmlth@t-hamel.fr">hmlth@t-hamel.fr</a><br>
> To: <a href="mailto:abauer@magix.net">abauer@magix.net</a><br>
> Cc: <a href="mailto:gluster-users@gluster.org">gluster-users@gluster.org</a><br>
> Sent: Monday, September 28, 2015 7:40:52 AM<br>
> Subject: Re: [Gluster-users] Tuning for small files<br>
><br>
> I'm also quite interested by small files performances optimization, but<br>
> I'm a bit confused about the best option between 3.6/3.7.<br>
><br>
> Ben Turner was saying that 3.6 might give the best performances:<br>
> <a href="http://www.gluster.org/pipermail/gluster-users/2015-September/023733.html" rel="noreferrer" target="_blank">http://www.gluster.org/pipermail/gluster-users/2015-September/023733.html</a><br>
><br>
> What kind of gain is expected (with consistent-metadata) if this<br>
> regression is solved?<br>
<br>
Just to be clear, the issue I am talking about is metadata only(think ls -l or file browsing). It doesn't affect small file perf(well not that much, I'm sure a little, but I have never quantified it), with server and client event threads set to 4 + lookup optimize I see between a 200-300% gain on my systems on 3.7 vs 3.6 builds. If I needed fast metadata I would go with 3.6, if I need fast smallfile I would go with 3.7. If I needed both I would pick the less of the two evils and go with that one and upgrade when the fix is released.<br>
<br>
-b<br>
<br>
<br>
><br>
> I tried 3.6.5 (last version for debian jessie), and it's a bit better<br>
> than 3.7.4 but not by much (10-15%).<br>
><br>
> I was also wondering if there is recommendations for the underlying file<br>
> system of the bricks (xfs, ext4, tuning...).<br>
><br>
><br>
> Regards<br>
><br>
> Thomas HAMEL<br>
><br>
> On 2015-09-28 12:04, André Bauer wrote:<br>
> > If you're not already on Glusterfs 3.7.x i would recommend an update<br>
> > first.<br>
> ><br>
> > Am 25.09.2015 um 17:49 schrieb Thibault Godouet:<br>
> >> Hi,<br>
> >><br>
> >> There are quite a few tuning parameters for Gluster (as seen in<br>
> >> Gluster<br>
> >> volume XYZ get all), but I didn't find much documentation on those.<br>
> >> Some people do seem to set at least some of them, so the knowledge<br>
> >> must<br>
> >> be somewhere...<br>
> >><br>
> >> Is there a good source of information to understand what they mean,<br>
> >> and<br>
> >> recommendation on how to set them to get a good small file<br>
> >> performance?<br>
> >><br>
> >> Basically what I'm trying to optimize is for svn operations (e.g. svn<br>
> >> checkout, or svn branch) on a replicated 2 x 1 volume (hosted on 2<br>
> >> VMs,<br>
> >> 16GB ram, 4 cores each, 10Gb/s network tested at full speed), using a<br>
> >> NFS mount which appears much faster than fuse in this case (but still<br>
> >> much slower than when served by a normal NFS server).<br>
> >> Any recommendation for such a setup?<br>
> >><br>
> >> Thanks,<br>
> >> Thibault.<br>
> >><br>
> >><br>
> >><br>
> >> _______________________________________________<br>
> >> Gluster-users mailing list<br>
> >> <a href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a><br>
> >> <a href="http://www.gluster.org/mailman/listinfo/gluster-users" rel="noreferrer" target="_blank">http://www.gluster.org/mailman/listinfo/gluster-users</a><br>
> >><br>
> ><br>
> ><br>
> > --<br>
> > Mit freundlichen Grüßen<br>
> > André Bauer<br>
> ><br>
> > MAGIX Software GmbH<br>
> > André Bauer<br>
> > Administrator<br>
> > August-Bebel-Straße 48<br>
> > 01219 Dresden<br>
> > GERMANY<br>
> ><br>
> > tel.: 0351 41884875<br>
> > e-mail: <a href="mailto:abauer@magix.net">abauer@magix.net</a><br>
> > <a href="mailto:abauer@magix.net">abauer@magix.net</a> <mailto:<a href="mailto:Email">Email</a>><br>
> > <a href="http://www.magix.com" rel="noreferrer" target="_blank">www.magix.com</a> <<a href="http://www.magix.com/" rel="noreferrer" target="_blank">http://www.magix.com/</a>><br>
> ><br>
> ><br>
> > Geschäftsführer | Managing Directors: Dr. Arnd Schröder, Michael Keith<br>
> > Amtsgericht | Commercial Register: Berlin Charlottenburg, HRB 127205<br>
> ><br>
> > Find us on:<br>
> ><br>
> > <<a href="http://www.facebook.com/MAGIX" rel="noreferrer" target="_blank">http://www.facebook.com/MAGIX</a>> <<a href="http://www.twitter.com/magix_de" rel="noreferrer" target="_blank">http://www.twitter.com/magix_de</a>><br>
> > <<a href="http://www.youtube.com/wwwmagixcom" rel="noreferrer" target="_blank">http://www.youtube.com/wwwmagixcom</a>> <<a href="http://www.magixmagazin.de" rel="noreferrer" target="_blank">http://www.magixmagazin.de</a>><br>
> > ----------------------------------------------------------------------<br>
> > The information in this email is intended only for the addressee named<br>
> > above. Access to this email by anyone else is unauthorized. If you are<br>
> > not the intended recipient of this message any disclosure, copying,<br>
> > distribution or any action taken in reliance on it is prohibited and<br>
> > may be unlawful. MAGIX does not warrant that any attachments are free<br>
> > from viruses or other defects and accepts no liability for any losses<br>
> > resulting from infected email transmissions. Please note that any<br>
> > views expressed in this email may be those of the originator and do> > Gluster-users mailing list<br>
> > <a href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a><br>
> > <a href="http://www.gluster.org/mailman/listinfo/gluster-users" rel="noreferrer" target="_blank">http://www.gluster.org/mailman/listinfo/gluster-users</a><br>
><br>
> _______________________________________________<br>
> Gluster-users mailing list<br>
> <a href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a><br>
> <a href="http://www.gluster.org/mailman/listinfo/gluster-users" rel="noreferrer" target="_blank">http://www.gluster.org/mailman/listinfo/gluster-users</a><br>
_______________________________________________<br>
Gluster-users mailing list<br>
<a href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a><br>
<a href="http://www.gluster.org/mailman/listinfo/gluster-users" rel="noreferrer" target="_blank">http://www.gluster.org/mailman/listinfo/gluster-users</a></div>