<html><head>
<meta http-equiv="Content-Type" content="text/html; charset=Windows-1252">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Thanks a lot for looking into this.<br>
    <br>
      Toni<br>
    <br>
    <br>
    <div class="moz-cite-prefix">On 10/10/2015 05:47 AM, Seo, Sangmin
      wrote:<br>
    </div>
    <blockquote cite="mid:AA4897B5-4779-4EEF-B3C0-BDE2BC87DFC5@anl.gov" type="cite">
      
      <div class="">Hi Toni,</div>
      <div class=""><br class="">
      </div>
      <div class="">We figured out the reason (the spawned processes
        hang because when the "hosts" info keys for two commands have
        the same value (e.g., localhost) two processes get the same
        rank.), but it seems it will take some time to fix this. Pavan
        will rewrite the problematic part in hydra.</div>
      <div class=""><br class="">
      </div>
      <div class="">I’ve created a ticket (<a moz-do-not-send="true" href="https://trac.mpich.org/projects/mpich/ticket/2308" class="">https://trac.mpich.org/projects/mpich/ticket/2308</a>)
        for this issue and cc’ed you in the ticket.</div>
      <div class=""><br class="">
      </div>
      <div class="">Regards,</div>
      <div class="">Sangmin</div>
      <div class=""><br class="">
      </div>
      <br class="">
      <div>
        <blockquote type="cite" class="">
          <div class="">On Oct 8, 2015, at 10:17 AM, Antonio J. Peña
            <<a moz-do-not-send="true" href="mailto:antonio.pena@bsc.es" class="">antonio.pena@bsc.es</a>>
            wrote:</div>
          <br class="Apple-interchange-newline">
          <div class=""><br class="">
            Thanks a lot Sangmin!<br class="">
            <br class="">
            <br class="">
            On 10/08/2015 03:43 PM, Seo, Sangmin wrote:<br class="">
            <blockquote type="cite" class="">Hi Toni,<br class="">
              <br class="">
              I confirmed the problem and am looking into it. I will get
              back to you after I resolve it.<br class="">
              <br class="">
              Regards,<br class="">
              Sangmin<br class="">
              <br class="">
              <br class="">
              <blockquote type="cite" class="">On Oct 6, 2015, at 7:23
                AM, Antonio J. Peña <<a moz-do-not-send="true" href="mailto:antonio.pena@bsc.es" class="">antonio.pena@bsc.es</a>>
                wrote:<br class="">
                <br class="">
                <br class="">
                Right, but I do need to use the hosts key to get control
                of where the processes are spawned. Any clue?<br class="">
                <br class="">
                <br class="">
                On 10/02/2015 07:32 PM, Thakur, Rajeev wrote:<br class="">
                <blockquote type="cite" class="">Looks like it is
                  related to the hosts key. Works without that.<br class="">
                  <br class="">
                  Rajeev<br class="">
                  <br class="">
                  <blockquote type="cite" class="">On Oct 2, 2015, at
                    10:34 AM, Antonio J. Peña <<a moz-do-not-send="true" href="mailto:antonio.pena@bsc.es" class=""><a class="moz-txt-link-abbreviated" href="mailto:antonio.pena@bsc.es">antonio.pena@bsc.es</a></a>>
                    wrote:<br class="">
                    <br class="">
                    Hi folks!<br class="">
                    <br class="">
                    I've been facing an issue with
                    MPI_Comm_spawn_multiple when all the<br class="">
                    following conditions are met:<br class="">
                    <br class="">
                     1. 'count' is > 1<br class="">
                     2. Any value of 'array_of_maxprocs' is > 1<br class="">
                     3. Setting the "hosts" info key<br class="">
                    <br class="">
                    What I'm finding is that the spawned processes hang
                    at MPI_Init. Find<br class="">
                    attached a test case. You can reproduce my issue
                    with -np 1.<br class="">
                    <br class="">
                    Is there anything I'm doing wrong? I'm using MPICH
                    master at 1e6c4d8. I<br class="">
                    haven't found a pending ticket, xfail test, or test
                    case covering this<br class="">
                    situation.<br class="">
                    <br class="">
                    All the best,<br class="">
                      Toni<br class="">
                    <br class="">
                    [P.S.: It still feels weird to be on the other side
                    of <a moz-do-not-send="true" href="mailto:discuss@mpich.org" class="">
                      discuss@mpich.org</a>]<br class="">
                    <br class="">
                    -- <br class="">
                    Antonio J. Peña<br class="">
                    Senior Researcher<br class="">
                    Barcelona Supercomputing Center<br class="">
                    <br class="">
                    <br class="">
                    <br class="">
                    WARNING / LEGAL TEXT: This message is intended only
                    for the use of the<br class="">
                    individual or entity to which it is addressed and
                    may contain<br class="">
                    information which is privileged, confidential,
                    proprietary, or exempt<br class="">
                    from disclosure under applicable law. If you are not
                    the intended<br class="">
                    recipient or the person responsible for delivering
                    the message to the<br class="">
                    intended recipient, you are strictly prohibited from
                    disclosing,<br class="">
                    distributing, copying, or in any way using this
                    message. If you have<br class="">
                    received this communication in error, please notify
                    the sender and<br class="">
                    destroy and delete any copies you may have received.<br class="">
                    <br class="">
                    <a moz-do-not-send="true" href="http://www.bsc.es/disclaimer%3Cspawn_multiple.c%3E__" class="">http://www.bsc.es/disclaimer<spawn_multiple.c>__</a>_____________________________________________<br class="">
                    discuss mailing list     <a moz-do-not-send="true" href="mailto:discuss@mpich.org" class="">discuss@mpich.org</a><br class="">
                    To manage subscription options or unsubscribe:<br class="">
                    <a moz-do-not-send="true" href="https://lists.mpich.org/mailman/listinfo/discuss" class="">https://lists.mpich.org/mailman/listinfo/discuss</a><br class="">
                  </blockquote>
                  _______________________________________________<br class="">
                  discuss mailing list     <a moz-do-not-send="true" href="mailto:discuss@mpich.org" class="">discuss@mpich.org</a><br class="">
                  To manage subscription options or unsubscribe:<br class="">
                  <a moz-do-not-send="true" href="https://lists.mpich.org/mailman/listinfo/discuss" class="">https://lists.mpich.org/mailman/listinfo/discuss</a><br class="">
                </blockquote>
                -- <br class="">
                Antonio J. Peña<br class="">
                Senior Researcher<br class="">
                Barcelona Supercomputing Center<br class="">
                <br class="">
                <br class="">
                WARNING / LEGAL TEXT: This message is intended only for
                the use of the<br class="">
                individual or entity to which it is addressed and may
                contain<br class="">
                information which is privileged, confidential,
                proprietary, or exempt<br class="">
                from disclosure under applicable law. If you are not the
                intended<br class="">
                recipient or the person responsible for delivering the
                message to the<br class="">
                intended recipient, you are strictly prohibited from
                disclosing,<br class="">
                distributing, copying, or in any way using this message.
                If you have<br class="">
                received this communication in error, please notify the
                sender and<br class="">
                destroy and delete any copies you may have received.<br class="">
                <br class="">
                <a moz-do-not-send="true" href="http://www.bsc.es/disclaimer" class="">http://www.bsc.es/disclaimer</a><br class="">
                _______________________________________________<br class="">
                discuss mailing list     <a class="moz-txt-link-abbreviated" href="mailto:discuss@mpich.org">discuss@mpich.org</a><br class="">
                To manage subscription options or unsubscribe:<br class="">
                <a class="moz-txt-link-freetext" href="https://lists.mpich.org/mailman/listinfo/discuss">https://lists.mpich.org/mailman/listinfo/discuss</a><br class="">
              </blockquote>
              _______________________________________________<br class="">
              discuss mailing list     <a moz-do-not-send="true" href="mailto:discuss@mpich.org" class="">discuss@mpich.org</a><br class="">
              To manage subscription options or unsubscribe:<br class="">
              <a moz-do-not-send="true" href="https://lists.mpich.org/mailman/listinfo/discuss" class="">https://lists.mpich.org/mailman/listinfo/discuss</a><br class="">
            </blockquote>
            <br class="">
            -- <br class="">
            Antonio J. Peña<br class="">
            Senior Researcher<br class="">
            Barcelona Supercomputing Center<br class="">
            <br class="">
            <br class="">
            WARNING / LEGAL TEXT: This message is intended only for the
            use of the<br class="">
            individual or entity to which it is addressed and may
            contain<br class="">
            information which is privileged, confidential, proprietary,
            or exempt<br class="">
            from disclosure under applicable law. If you are not the
            intended<br class="">
            recipient or the person responsible for delivering the
            message to the<br class="">
            intended recipient, you are strictly prohibited from
            disclosing,<br class="">
            distributing, copying, or in any way using this message. If
            you have<br class="">
            received this communication in error, please notify the
            sender and<br class="">
            destroy and delete any copies you may have received.<br class="">
            <br class="">
            <a moz-do-not-send="true" href="http://www.bsc.es/disclaimer" class="">http://www.bsc.es/disclaimer</a><br class="">
            _______________________________________________<br class="">
            discuss mailing list     <a class="moz-txt-link-abbreviated" href="mailto:discuss@mpich.org">discuss@mpich.org</a><br class="">
            To manage subscription options or unsubscribe:<br class="">
            <a class="moz-txt-link-freetext" href="https://lists.mpich.org/mailman/listinfo/discuss">https://lists.mpich.org/mailman/listinfo/discuss</a></div>
        </blockquote>
      </div>
      <br class="">
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
discuss mailing list     <a class="moz-txt-link-abbreviated" href="mailto:discuss@mpich.org">discuss@mpich.org</a>
To manage subscription options or unsubscribe:
<a class="moz-txt-link-freetext" href="https://lists.mpich.org/mailman/listinfo/discuss">https://lists.mpich.org/mailman/listinfo/discuss</a></pre>
    </blockquote>
    <br>
    <pre class="moz-signature" cols="72">-- 
Antonio J. Peña
Senior Researcher
Barcelona Supercomputing Center</pre>
  
<br>
<br>
WARNING / LEGAL TEXT: This message is intended only for the use of the
individual or entity to which it is addressed and may contain
information which is privileged, confidential, proprietary, or exempt
from disclosure under applicable law. If you are not the intended
recipient or the person responsible for delivering the message to the
intended recipient, you are strictly prohibited from disclosing,
distributing, copying, or in any way using this message. If you have
received this communication in error, please notify the sender and
destroy and delete any copies you may have received.
<br><br>
<a href="http://www.bsc.es/disclaimer.htm">http://www.bsc.es/disclaimer</a>
<br>
</body>
</html>