<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Sep 6, 2017 at 2:47 PM, Gianluca Cecchi <span dir="ltr"><<a href="mailto:gianluca.cecchi@gmail.com" target="_blank">gianluca.cecchi@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr">Hello,<div>I was talking with a guy expert in VMware and discussing performance of VMs in respect of virtual cpus assigned to them in relation with mapping with the real hw of the hypervisor underneath.</div><div><br></div><div>One of the topics was numa usage and its overheads in case of a "too" big VM, in terms of both number of vcpus and memory amount.</div><div>Eg: </div><div>suppose host has 2 intel based sockets, with 6 cores and HT enabled and has 96Gb of ram (distributed 48+48 between the 2 processors)</div><div>suppose I configure a VM with 16 vcpus (2:4:2): would be the mapping respected at physical level or only a sort of "hint" for the hypervisor?</div><div>Can I say that it would perform better if I configure it 12 vcpus and mapping 1:6:2, because it can stay all inside one cpu?</div></div></blockquote><div><br></div><div>Hard to say without relationship to the workload. You are losing 4 vCPUs - perhaps those can be used for something (the OS) while the rest of them could be used by the application, for example?</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div><br></div><div>And what if I define a VM with 52Gb of ram? Can I say that it would perform in general better if I try to get it all in one cpu related memory slots (eg not more than 48Gb in my example)?</div><div><br></div></div></blockquote><div>Hard to say without relationship to the workload - will it need all the memory? Will it be accessing all of it, in random order?</div><div><br></div><div>If you've maxed out one node, you just need more memory from the other node. </div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div></div><div>Are there any documents going more deeply in these sort of considerations?</div><div><br></div></div></blockquote><div><br></div><div>It is so workload dependent that there will not be a one size fit all.</div><div><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 dir="ltr"><div></div><div>Also, if one goes and sizes so that the biggest VM is able to all-stay inside one cpu-memory, does it make sense to say that it will perform better in this scenario a cluster composed by 4 nodes, each one with 1 socket and 48Gb of memory instead of a cluster of 2 nodes, each one with 2 sockets and 96Gb of ram?</div></div></blockquote><div><br></div><div>You could have used affinity.</div><div><br></div><div>See [1] for some details on Redis. Note that IO (specifically network) is just as important - and its impact is much more profound.</div><div>Y.<br></div><div><br></div><div>[1] <a href="https://redis.io/topics/benchmarks">https://redis.io/topics/benchmarks</a></div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div><br></div><div>Hope I have clarified my questions/doubts.</div><div><br></div><div><br></div><div>Thanks in advance for any insight,</div><div>Gianluca </div></div>
<br>______________________________<wbr>_________________<br>
Users mailing list<br>
<a href="mailto:Users@ovirt.org">Users@ovirt.org</a><br>
<a href="http://lists.ovirt.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://lists.ovirt.org/<wbr>mailman/listinfo/users</a><br>
<br></blockquote></div><br></div></div>