<div dir="ltr"><br><div class="gmail_extra"><br><br><div class="gmail_quote">On Thu, Jun 6, 2013 at 11:21 AM, Brian Lamar <span dir="ltr"><<a href="mailto:brian.lamar@rackspace.com" target="_blank">brian.lamar@rackspace.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
<div style="word-wrap:break-word">
OK, I'll bite.
<div><br>
</div>
<div>Can you elaborate on "work at scale"? Do you have numbers you'd like to see? Are those numbers for a single-node installation? I've spun up tons of 500 VM environments and it doesn't seem broken to me.</div>
</div></blockquote><div><br></div><div style>Keeping 500 VMS running is very different then trying to start 500 VMs simultaneously. </div><div><br></div><div style>I don't have specific numbers I would like to see, but faster is always better. Yes these numbers are for a single-node installation. The POST request to boot 100 VMs takes 8.4 seconds! That alone seems too slow, an 8.4 second REST call. And the VMs take about 160 seconds to boot.</div>
<div><br></div><div>As far as I can tell things are not 'broken' when running 100's of VMs in devstack. Just some things become very very slow, for example RPC calls to nova-network begin to time out when trying to boot 100 VMs.</div>
<div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div style="word-wrap:break-word"><div>
</div>
<div>This hopefully doesn't come across as hostile, as I'm truly curious what your thoughts are on 'scale'.</div>
<div><br>
</div>
<div>Brian</div>
<div><br>
</div>
<div><br>
</div>
<div>
<div>
<div>On Jun 6, 2013, at 1:22 PM, Joe Gordon <<a href="mailto:joe.gordon0@gmail.com" target="_blank">joe.gordon0@gmail.com</a>></div>
<div> wrote:</div>
<br>
<blockquote type="cite"><div><div>
<div dir="ltr">
<div>TL;DR<br>
</div>
<div>Using this patch (<a href="https://review.openstack.org/#/c/31876/" target="_blank">https://review.openstack.org/#/c/31876/</a>) set 'VIRT_DRIVER=fake' in your devstack rc file. And now you can try spawning up as many VMs as you want.<br>
</div>
<div><br>
</div>
<div><br>
</div>
<div>I have just proposed a patch to devstack to use nova's fake virt driver. This can be used to see how OpenStack runs when trying to launch hundreds or thousands of VMs at once. Turns out it runs really really slowly.</div>
<div><br>
</div>
<div>I am sending out this to the mailing list to ask for help testing (and filing bugs) using this, so that we as community can be more confident things will work at scale (defined as number of VMs here).</div>
<div><br>
</div>
<div><br>
</div>
<div>best,</div>
<div>Joe</div>
</div></div></div>
_______________________________________________<br>
OpenStack-dev mailing list<br>
<a href="mailto:OpenStack-dev@lists.openstack.org" target="_blank">OpenStack-dev@lists.openstack.org</a><br>
<a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a><br>
</blockquote>
</div>
<br>
</div>
</div>
<br>_______________________________________________<br>
OpenStack-dev mailing list<br>
<a href="mailto:OpenStack-dev@lists.openstack.org" target="_blank">OpenStack-dev@lists.openstack.org</a><br>
<a href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev" target="_blank">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a><br>
<br></blockquote></div><br></div></div>