<div dir="ltr">Thanks for sharing Julio. It seems like overall the numbers look better with QEMU 4.<div><br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Wed, May 29, 2019 at 8:36 AM Montes, Julio <<a href="mailto:julio.montes@intel.com">julio.montes@intel.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div style="text-align:left;direction:ltr">
<div>Hi</div>
<div><br>
</div>
<div>In order to reduce the noise, I re-ran the tests in a new no-GUI system. </div>
<div>In the case of QEMU4+PVH the boot time is a little bit better but the memory</div>
<div>footprint is bigger than qemu4 and qemu-lite. Here the results</div>
<div><br>
</div>
<div><br>
</div>
<div><img src="cid:16b053ab392c1e7e07c2"><br>
</div>
<div><br>
</div>
<div><img src="cid:16b053ab392c1f5f7fd1"><br>
</div>
<div><br>
</div>
<div>On Wed, 2019-05-29 at 11:07 +0200, Stefano Garzarella wrote:</div>
<blockquote type="cite" style="margin:0px 0px 0px 0.8ex;border-left:2px solid rgb(114,159,207);padding-left:1ex">
<div>On Tue, May 28, 2019 at 07:21:38PM +0000, Montes, Julio wrote:</div>
<blockquote type="cite" style="margin:0px 0px 0px 0.8ex;border-left:2px solid rgb(114,159,207);padding-left:1ex">
<div>On Tue, 2019-05-28 at 12:10 -0700, Maran Wilson wrote:</div>
<blockquote type="cite" style="margin:0px 0px 0px 0.8ex;border-left:2px solid rgb(114,159,207);padding-left:1ex">
<div>On 5/28/2019 11:54 AM, Whaley, Graham wrote:</div>
<blockquote type="cite" style="margin:0px 0px 0px 0.8ex;border-left:2px solid rgb(114,159,207);padding-left:1ex">
<blockquote type="cite" style="margin:0px 0px 0px 0.8ex;border-left:2px solid rgb(114,159,207);padding-left:1ex">
<div>-----Original Message-----</div>
<div>From: Dr. David Alan Gilbert [mailto:<a href="mailto:dgilbert@redhat.com" target="_blank">dgilbert@redhat.com</a>]</div>
<div>Sent: Tuesday, May 28, 2019 7:40 PM</div>
<div>To: Montes, Julio <<a href="mailto:julio.montes@intel.com" target="_blank">julio.montes@intel.com</a>>;
<a href="mailto:pbonzini@redhat.com" target="_blank">pbonzini@redhat.com</a></div>
<div>Cc: <a href="mailto:kata-dev@lists.katacontainers.io" target="_blank">kata-dev@lists.katacontainers.io</a></div>
<div>Subject: Re: [kata-dev] QEMU 4 and QEMU-lite</div>
<div><br>
</div>
<div>* Montes, Julio (<a href="mailto:julio.montes@intel.com" target="_blank">julio.montes@intel.com</a>) wrote:</div>
<blockquote type="cite" style="margin:0px 0px 0px 0.8ex;border-left:2px solid rgb(114,159,207);padding-left:1ex">
<div>On Tue, 2019-05-28 at 18:19 +0100, Dr. David Alan Gilbert</div>
<div>wrote:</div>
<blockquote type="cite" style="margin:0px 0px 0px 0.8ex;border-left:2px solid rgb(114,159,207);padding-left:1ex">
<div>* Montes, Julio (<a href="mailto:julio.montes@intel.com" target="_blank">julio.montes@intel.com</a>) wrote:</div>
<blockquote type="cite" style="margin:0px 0px 0px 0.8ex;border-left:2px solid rgb(114,159,207);padding-left:1ex">
<div>Hi kata-folks</div>
<div><br>
</div>
<div>One of the topics in today's agent was QEMU 4 and the work</div>
<div>that we</div>
<div>are doing to support it in Kata Containers.</div>
<div>I ran the boot time and memory footprint metrics in my</div>
<div>workstation</div>
<div>to compare QEMU 4 vs QEMU-lite (2.11).</div>
<div>Here the results.</div>
<div><br>
</div>
<div>Feel free to comment and raise your concerns about QEMU 4.</div>
</blockquote>
<div><br>
</div>
<div>So that looks pretty nice for QEMU 4; at least as fast and in</div>
<div>the noise on the memory usage?</div>
</blockquote>
<div><br>
</div>
<div>yeah - IMO QEMU 4 has a good performance and new features, it's</div>
<div>a good</div>
<div>hypervisor for Kata Containers.</div>
</blockquote>
<div><br>
</div>
<div>Great! Please keep running these tests to make sure we don't</div>
<div>regress in</div>
<div>future versions (especially on our soft-freezes etc so we can</div>
<div>spot</div>
<div>them before the next release.</div>
</blockquote>
<div><br>
</div>
<div>We do have a 'metrics CI' running on Kata, that in theory would</div>
<div>detect any major (roughly</div>
<div>+/- 5%) shifts - but, that would only happen when Kata updates the</div>
<div>version of Qemu it is tracking...</div>
<div>And, yes, we will almost definitely have to tweak that CI to</div>
<div>account for the new numbers when we land</div>
<div>The qemu4 PR ...</div>
</blockquote>
<div><br>
</div>
<div>Just curious about whether you are providing the uncompressed kernel </div>
<div>image in these Qemu 4 tests or not. Qemu 4.0 supports booting the </div>
<div>uncompressed kernel binary in order to reduce boot time (if the</div>
<div>kernel </div>
</blockquote>
<div><br>
</div>
<div>\o/ interesting I will try, thanks.</div>
<div><br>
</div>
</blockquote>
<div><br>
</div>
<div>As Maran, I'm also curious about a comparison with the new QEMU 4.0</div>
<div>feature to boot uncompressed kernel binary.</div>
<div><br>
</div>
<div>Here you can find some useful information (I hope :)):</div>
<div><a href="https://gist.github.com/stefano-garzarella/7b7e17e75add20abd1c42fb496cc6504" target="_blank">https://gist.github.com/stefano-garzarella/7b7e17e75add20abd1c42fb496cc6504</a></div>
<div><br>
</div>
<div>Thanks,</div>
<div>Stefano</div>
</blockquote>
</div>
_______________________________________________<br>
kata-dev mailing list<br>
<a href="mailto:kata-dev@lists.katacontainers.io" target="_blank">kata-dev@lists.katacontainers.io</a><br>
<a href="http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev" rel="noreferrer" target="_blank">http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev</a><br>
</blockquote></div>