<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">Hi Mike and Zane,<br>
      <br>
      Le 27/09/2013 15:58, Mike Spreitzer a écrit :<br>
    </div>
    <blockquote
cite="mid:OF7C2021B1.BDBBD7E2-ON85257BF3.004C1B84-85257BF3.004CBFC3@us.ibm.com"
      type="cite">
      <meta http-equiv="Content-Type" content="text/html;
        charset=ISO-8859-1">
      <tt><font size="2">Zane Bitter <a class="moz-txt-link-rfc2396E" href="mailto:zbitter@redhat.com"><zbitter@redhat.com></a> wrote on
          09/27/2013
          08:24:49 AM:<br>
          <br>
          > Your diagrams clearly show scheduling happening in a
          separate stage
          to <br>
          > (infrastructure) orchestration, which is to say that at
          the point
          where <br>
          > resources are scheduled, their actual creation is in the
          *future*.<br>
          > <br>
          > I am not a Climate expert, but it seems to me that they
          have a <br>
          > near-identical problem to solve: how do they integrate
          with Heat such
          <br>
          > that somebody who has reserved resources in the past can
          actually
          create <br>
          > them (a) as part of a Heat stack or (b) as standalone
          resources, at
          the <br>
          > user's option. IMO OpenStack should solve this problem
          only once.</font></tt>
      <br>
      <br>
      <tt><font size="2">If I understand correctly, what Climate adds to
          the
          party is planning allocations to happen at some specific time
          in the non-immediate
          future.  A holistic infrastructure scheduler is planning
          allocations
          to happen just as soon as we can get the plans through the
          relevant code
          path, which is why I describe it as "now".</font></tt>
      <br>
      <tt><font size="2"><br>
        </font></tt></blockquote>
    <br>
    <tt><font size="2">Cl<font size="2">imate <font size="2">is wide-s<font
              size="2">cope<font size="2">d aiming <font size="2">to <font
                    size="2">e<font size="2">xc<font size="2">lusively
                        reserve any kind of resou<font size="2">rces by
                          a certain time. This gene<font size="2">ric
                            sentence doesn't mean Climate ca<font
                              size="2">n't <font size="2">schedule
                                things <font size="2">'now<font
                                    size="2">'<font size="2"> : you can
                                      ask for an immediate lease
                                      (starting <font size="2">'now<font
                                          size="2">')</font></font> and
                                      <font size="2">you<font size="2">
                                          will <font size="2">get the
                                            resources as of now<font
                                              size="2">.<font size="2"><br>
                                                <font size="2"><br>
                                                  <font size="2"><font
                                                      size="2">Climate <font
                                                        size="2">team is
                                                        actually split <font
                                                          size="2">i<font
                                                          size="2">nto t<font
                                                          size="2">wo
                                                          different
                                                          teams, one
                                                          focusin<font
                                                          size="2">g on
                                                          hardware
                                                          procurement
                                                          and one
                                                          focusing of
                                                          virtual p<font
                                                          size="2">rocurement.
                                                          I can't spe<font
                                                          size="2">ak on
                                                          <font size="2">behalf
                                                          of the <font
                                                          size="2">'Climate
                                                          Virtual' team</font></font></font></font></font></font></font></font></font></font></font>,
                                                  but I would bet <font
                                                    size="2">scheduling
                                                    <font size="2">an
                                                      Heat stack or a<font
                                                        size="2"> S<font
                                                          size="2">avan<font
                                                          size="2">na c<font
                                                          size="2">luster
                                                          will require
                                                          some kind of
                                                          holis<font
                                                          size="2">tic D<font
                                                          size="2">SL,
                                                          indeed.<br>
                                                          <br>
                                                          <font size="2">From
                                                          the 'Climate P<font
                                                          size="2">h<font
                                                          size="2">ysical</font></font></font>'
                                                          <font size="2">POV,
                                                          that could
                                                          even be
                                                          necessary, but<font
                                                          size="2"> yet<font
                                                          size="2">
                                                          unclear a<font
                                                          size="2">t the
                                                          moment.<br>
                                                          <br>
                                                          <font size="2">-Sylvain</font><br>
                                                          </font></font></font></font></font></font></font></font></font></font></font></font><br>
                                                </font></font></font></font></font></font></font></font></font></font></font></font></font></font></font></font></font></font></font></font></font></font></tt>
    <blockquote
cite="mid:OF7C2021B1.BDBBD7E2-ON85257BF3.004C1B84-85257BF3.004CBFC3@us.ibm.com"
      type="cite"><tt><font size="2">
          <br>
          > If I understood your remarks correctly, we agree that
          there is no
          <br>
          > (known) reason that the scheduling has to occur in the
          middle of <br>
          > orchestration (which would have implied that it needed to
          be <br>
          > incorporated in some sense into Heat).</font></tt>
      <br>
      <br>
      <tt><font size="2">If you agree that by orchestration you meant
          specifically
          infrastructure orchestration then we are agreed.  If software
          orchestration
          is also in the picture then I also agree that holistic
          infrastructure scheduling
          does not *have to* go in between software orchestration and
          infrastructure
          orchestration --- but I think that's a pretty good place for
          it.</font></tt>
      <br>
      <br>
      <tt><font size="2"><br>
          > Right, so what I'm saying is that if all those things are
          _stated_
          in <br>
          > the input then there's no need to run the orchestration
          engine to
          find <br>
          > out what they'll be; they're already stated.<br>
        </font></tt>
      <br>
      <tt><font size="2">Yep.</font></tt>
      <br>
      <br>
      <tt><font size="2">Thanks,</font></tt>
      <br>
      <tt><font size="2">Mike</font></tt>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
OpenStack-dev mailing list
<a class="moz-txt-link-abbreviated" href="mailto:OpenStack-dev@lists.openstack.org">OpenStack-dev@lists.openstack.org</a>
<a class="moz-txt-link-freetext" href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a>
</pre>
    </blockquote>
    <br>
  </body>
</html>