<div dir="ltr"><div dir="ltr"><div class="gmail_default" style="font-size:small">Hi Matt,</div><div class="gmail_default" style="font-size:small"><br></div><div class="gmail_default" style="font-size:small">Thanks for looking into this,</div><div class="gmail_default" style="font-size:small"><br></div><br><div class="gmail_quote"><div dir="ltr">On Wed, Dec 5, 2018 at 10:27 PM Matt Riedemann <<a href="mailto:mriedemos@gmail.com">mriedemos@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">Belmiro/Surya,<br>
<br>
I'm trying to follow up on something Belmiro mentioned at the summit <br>
before I forget about it.<br>
<br>
CERN sets this value low:<br>
<br>
<a href="https://docs.openstack.org/nova/latest/configuration/config.html#scheduler.max_placement_results" rel="noreferrer" target="_blank">https://docs.openstack.org/nova/latest/configuration/config.html#scheduler.max_placement_results</a><br>
<br>
And as a result, when disabling nova-computes during maintenance, you <br>
can fail during scheduling because placement only returns resource <br>
providers for disabled computes.<br>
<br>
I believe Dan and I kicked around some ideas on how we could deal with <br>
this, like either via a periodic in the compute service or when the <br>
compute service is disabled in the API, we would set the 'reserved' <br>
inventory value equal to the total to take those computes out of <br>
scheduling. </blockquote><div><br></div><div><div class="gmail_default" style="font-size:small">Just read the discussion on the channel and saw there were a couple of approaches proposed like traits and neg-aggregates in addition to the above two.</div></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">I think Belmiro said this is what CERN is doing today as a <br>
workaround?<br>
<br></blockquote><div><br></div><div><div class="gmail_default" style="font-size:small">As far as I know we don't have it in PROD, I will let Belmiro confirm this anyways</div><br></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">
For the latter solution, I don't know if we'd proxy that change directly <br>
from nova-api to placement, or make an RPC cast/call to nova-compute to <br>
do it, but that's an implementation detail.<br>
<br>
I mostly just want to make sure we get a bug reported for this so we <br>
don't lose track of it. Can one of you open a bug with your scenario and <br>
current workaround?<br><br></blockquote><div><br></div><div class="gmail_default" style="font-size:small">We have already filed a bug for this: <a href="https://bugs.launchpad.net/nova/+bug/1805984">https://bugs.launchpad.net/nova/+bug/1805984</a>. Will add the workaround we have into the description.</div></div><div><br></div>--<span class="gmail_default" style="font-size:small">----------</span><div dir="ltr" class="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div>Regards,<br></div><div>Surya.<br></div></div></div></div></div></div></div>