<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<div class="moz-cite-prefix">Hi Laurent,</div>
<div class="moz-cite-prefix"><br>
</div>
<div class="moz-cite-prefix">Thank you for your answer.<br>
</div>
<div class="moz-cite-prefix">I agree with you that without the
pinning the steal time can appear anytime. What is strange to me
that in openstack Kilo the steal is significantly smaller even
when there is some overcommit. So I am wondering where to look for
the difference?<br>
</div>
<div class="moz-cite-prefix"><br>
</div>
<div class="moz-cite-prefix">Kind regards,</div>
<div class="moz-cite-prefix">Laszlo<br>
</div>
<div class="moz-cite-prefix"><br>
</div>
<div class="moz-cite-prefix">On 11/12/20 5:01 PM, Laurent Dumont
wrote:<br>
</div>
<blockquote type="cite"
cite="mid:CAOAKi8zcoAsTH3BAAk_7fq4TJ0QnocCK30eNiEsjnYtcBj5YJA@mail.gmail.com">
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<div dir="ltr">Hi,
<div><br>
</div>
<div>Technically, I think that you always run the chance of
"steal" time if you don't pin CPUs. I'm not sure if Openstack
is "smart" enough to allocate CPUs that are not mapped to
anyone in sequence (and start to overcommit once it's
necessary. You might have 42 out of 48 free CPUs but I don't
think that means that Openstack will prevent two VM from
getting the same CPU scheduled (without CPU pinning).</div>
</div>
<br>
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Thu, Nov 12, 2020 at 9:09
AM Budai Laszlo <<a href="mailto:laszlo.budai@gmail.com"
moz-do-not-send="true">laszlo.budai@gmail.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>
<p>Hello all,</p>
<p>we are comparing the behavior of our queens openstack
with kilo. In queens we are observing an increase in the
steal time reported in the guest along with the increase
of the load averages. All this is happening while the host
is not overloaded, and reports 80+ idle time<br>
</p>
<p>Initially we have suspected that the overcommit might be
the reason of the steal, so we have migrated vms, and now
there are 42 vCPUs used out of the 48 pCPUs, but in the
guest we still observe the steal time.</p>
<p>with similar configuration in openstack kilo we see
smaller load, and almost no steal time at all.</p>
<p>what could be the reason of this steal time when there is
no CPU overcommit?</p>
<p>Thank you for any ideas.</p>
<p>Kind regards,<br>
Laszlo<br>
</p>
</div>
</blockquote>
</div>
</blockquote>
<p><br>
</p>
</body>
</html>