<div dir="ltr">try it, i bet that you will get better latency results with proper configured iscsitarget/initiator. <div><br></div><div>btw, freebsd 10 includes kernel based iscsi-target now. which works pretty good for me since some time, easy to setup and working performing well (zfs not to forget ;) )</div>
</div><div class="gmail_extra"><br><br><div class="gmail_quote">On Thu, Jan 9, 2014 at 9:20 AM, Markus Stockhausen <span dir="ltr"><<a href="mailto:stockhausen@collogia.de" target="_blank">stockhausen@collogia.de</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">> Von: Karli Sjöberg [<a href="mailto:Karli.Sjoberg@slu.se">Karli.Sjoberg@slu.se</a>]<br>
> Gesendet: Donnerstag, 9. Januar 2014 08:48<br>
> An: <a href="mailto:squadra@gmail.com">squadra@gmail.com</a><br>
> Cc: <a href="mailto:users@ovirt.org">users@ovirt.org</a>; Markus Stockhausen<br>
<div class="im">> Betreff: Re: [Users] Experience with low cost NFS-Storage as VM-Storage?<br>
><br>
</div><div class="im">> On Thu, 2014-01-09 at 08:35 +0100, squadra wrote:<br>
> Right, try multipathing with nfs :)<br>
><br>
> Yes, that´s what I meant, maybe could have been more clear about that,<br>
> sorry. Multipathing (and the load-balancing it brings) is what really<br>
> separates iSCSI from NFS.<br>
><br>
> What I´d be interested in knowing is at what breaking-point, not having<br>
> multipathing becomes an issue. I mean, we might not have such a big<br>
> VM-park, about 300-400 VMs. But so far running without multipathing<br>
> using good ole' NFS and no performance issues this far. Would be good to<br>
> know beforehand if we´re headed for a wall of some sorts, and about<br>
> "when" we´ll hit it...<br>
><br>
>/K<br>
<br>
</div>If that is really a concern for the initial question about a "low cost NFS<br>
solution" LACP on the NFS filer side will mitigate the bottleneck from<br>
too many hypervisors.<br>
<br>
My personal headache is the I/O performance of QEMU. More details here:<br>
<a href="http://lists.nongnu.org/archive/html/qemu-discuss/2013-12/msg00028.html" target="_blank">http://lists.nongnu.org/archive/html/qemu-discuss/2013-12/msg00028.html</a><br>
Or to make it short: Each I/O in a VM gets a penalty of 370us. That is much<br>
more than in ESX environments.<br>
<br>
I would be interested if this the same in ISCSI setups.<br>
<span class="HOEnZb"><font color="#888888"><br>
Markus<br>
</font></span></blockquote></div><br><br clear="all"><div><br></div>-- <br><pre>Sent from the Delta quadrant using Borg technology!</pre>
</div>