<div dir="ltr">Hi all,<div><br></div><div>Thank you for helping me on this error. Actually, real.exe is a portion of a very large weather model. It is very difficult to extract it or duplicate the error in a simple fortran code, since I am not sure where the problem is. From your discussion, I barely can understand them, in fact. Even I do not know what is "_get_addr". Is it related to MPI?</div><div><br></div><div>Best,</div><div>Zhifeng</div><div><br></div></div><br><div class="gmail_quote"><div dir="ltr">On Mon, Nov 19, 2018 at 10:06 AM Zhou, Hui via discuss <<a href="mailto:discuss@mpich.org">discuss@mpich.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">



<div style="word-wrap:break-word;line-break:after-white-space">
I suspect the error was something else. `real.exe: error: _get_addr: No such file or directory` does look like a MPICH error. In fact, I suspect mpirun probably is from other package by not hydra, then there is probably some mis-matched assumptions.
<div><br>
<div>
<div dir="auto" style="color:rgb(0,0,0);letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;word-wrap:break-word;line-break:after-white-space">
<div dir="auto" style="color:rgb(0,0,0);letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;word-wrap:break-word;line-break:after-white-space">
<div dir="auto" style="color:rgb(0,0,0);letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;word-wrap:break-word;line-break:after-white-space">
<div dir="auto" style="word-wrap:break-word;line-break:after-white-space">
<div style="color:rgb(0,0,0);font-family:Helvetica;font-size:12px;font-style:normal;font-variant-caps:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none">
— </div>
<div style="color:rgb(0,0,0);font-family:Helvetica;font-size:12px;font-style:normal;font-variant-caps:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none">
Hui Zhou</div>
<div style="color:rgb(0,0,0);font-family:Helvetica;font-size:12px;font-style:normal;font-variant-caps:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none">
T: 630-252-3430</div>
<div style="color:rgb(0,0,0);font-family:Helvetica;font-size:12px;font-style:normal;font-variant-caps:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none">
<br>
</div>
<div style="color:rgb(0,0,0);font-family:Helvetica;font-size:12px;font-style:normal;font-variant-caps:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none">
<br>
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
<div><br>
<blockquote type="cite">
<div>On Nov 17, 2018, at 12:31 PM, William Gropp <<a href="mailto:wgropp@illinois.edu" target="_blank">wgropp@illinois.edu</a>> wrote:</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
<div>
<div style="word-wrap:break-word">
I’m also at a loss to understand why "mpirun ./real.exe” would fail - if a -np xx argument is required, then mpirun should issue the error; if not, then a rational choice of default would be 1 process, and should behave as “mpirun -np 1 ./real.exe”.
<div><br>
</div>
<div>Of course, it should be mpiexec instead of mpirun, but that shouldn’t matter.</div>
<div><br>
</div>
<div>Bill</div>
<div><br>
<div>
<div style="letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;word-wrap:break-word;line-break:after-white-space">
<div style="letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;word-wrap:break-word;line-break:after-white-space">
<div style="letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;word-wrap:break-word;line-break:after-white-space">
<div style="word-wrap:break-word">
<div style="font-family:Helvetica;font-size:12px;font-style:normal;font-variant-caps:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px">
William Gropp<br>
Director and Chief Scientist, NCSA<br>
Thomas M. Siebel Chair in Computer Science<br>
University of Illinois Urbana-Champaign</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
<br>
<div>
<blockquote type="cite">
<div>On Nov 15, 2018, at 1:01 PM, Gropp, William D via discuss <<a href="mailto:discuss@mpich.org" target="_blank">discuss@mpich.org</a>> wrote:</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
<div>

<div style="word-wrap:break-word;line-break:after-white-space">
<div dir="auto">Yes, singleton is the ./real.exe below. At least that version should have worked.
<div dir="auto">Bill</div>
</div>
<div class="gmail_extra"><br>
<div class="gmail_quote">On Nov 14, 2018 2:00 PM, "Zhou, Hui" <<a href="mailto:zhouh@anl.gov" target="_blank">zhouh@anl.gov</a>> wrote:<br type="attribution">
</div>
</div>
<div>Hi Bill,
<div><br>
</div>
<div>To clarify,  a singleton refers to a process started directly — such as ./real.exe, right? For the case of `mpirun ./real.exe`, or even `mpirun -np 1 ./real.exe`, is not technically a singleton, correct?</div>
<div><br>
<div>
<div dir="auto" style="letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;word-wrap:break-word">
<div dir="auto" style="letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;word-wrap:break-word">
<div dir="auto" style="letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none;word-wrap:break-word">
<div dir="auto" style="word-wrap:break-word;line-break:after-white-space">
<div style="font-family:Helvetica;font-size:12px;font-style:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none">
— </div>
<div style="font-family:Helvetica;font-size:12px;font-style:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none">
Hui Zhou</div>
<div style="font-family:Helvetica;font-size:12px;font-style:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;text-decoration:none">
<br>
</div>
</div>
</div>
</div>
</div>
</div>
<div><br>
<blockquote type="cite">
<div>On Nov 13, 2018, at 9:30 PM, William Gropp via discuss <<a href="mailto:discuss@mpich.org" target="_blank">discuss@mpich.org</a>> wrote:</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
<div>
<div style="word-wrap:break-word">This is not quite true.  From the MPI standard:
<div><br>
</div>
<div>
<blockquote type="cite">
<div class="m_-9136802374877624834page" title="Page 428">
<div class="m_-9136802374877624834layoutArea">
<div class="m_-9136802374877624834column">
<p><span style="font-size:12pt;font-family:CMSS12">10.5.2 Singleton MPI</span><span style="font-size:12pt;font-family:CMTT12">_</span><span style="font-size:12pt;font-family:CMSS12">INIT</span></p>
</div>
</div>
</div>
</blockquote>
<blockquote type="cite">
<div class="m_-9136802374877624834page" title="Page 428">
<div class="m_-9136802374877624834layoutArea">
<div class="m_-9136802374877624834column">
<p><span style="font-family:CMR10;font-size:11pt">A high-quality implementation will allow any process (including those not started with a</span><span style="font-size:6pt;font-family:CMR6;vertical-align:-1pt"> </span><span style="font-size:11pt;font-family:CMR10">“parallel
 application” mechanism) to become an </span><span style="font-size:11pt;font-family:CMSS10">MPI
</span><span style="font-size:11pt;font-family:CMR10">process by calling
</span><span style="font-size:11pt;font-family:CMSS10">MPI</span><span style="font-size:11pt;font-family:CMTT10">_</span><span style="font-size:11pt;font-family:CMSS10">INIT</span><span style="font-size:11pt;font-family:CMR10">.
  </span></p>
</div>
</div>
</div>
</blockquote>
<div><br class="m_-9136802374877624834webkit-block-placeholder">
</div>
<div>There was a time when MPICH could do this, and there were features in the original PMI to support this.</div>
<div><br>
</div>
<div>Bill</div>
<div><br>
</div>
<div>
<div style="letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;word-wrap:break-word;line-break:after-white-space">
<div style="letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;word-wrap:break-word;line-break:after-white-space">
<div style="letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px;word-wrap:break-word;line-break:after-white-space">
<div style="word-wrap:break-word">
<div style="font-family:Helvetica;font-size:12px;font-style:normal;font-weight:normal;letter-spacing:normal;text-align:start;text-indent:0px;text-transform:none;white-space:normal;word-spacing:0px">
William Gropp<br>
Director and Chief Scientist, NCSA<br>
Thomas M. Siebel Chair in Computer Science<br>
University of Illinois Urbana-Champaign</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
<br class="m_-9136802374877624834Apple-interchange-newline">
</div>
<br>
<div>
<blockquote type="cite">
<div>On Nov 13, 2018, at 10:27 AM, Si, Min via discuss <<a href="mailto:discuss@mpich.org" target="_blank">discuss@mpich.org</a>> wrote:</div>
<br class="m_-9136802374877624834Apple-interchange-newline">
<div>
<div bgcolor="#FFFFFF">Hi Zhifeng,<br>
<br>
You have to always execute the MPI program by using mpirun or mpiexec with -np <number of processes>.<br>
<br>
At MPI_Init, the process tries to read information (e.g., MPI rank, number of processes) from the PMI server which is another process launched by the mpiexec command. This information is necessary for MPI library to initialize communication (e.g., connecting
 two processes).<br>
<br>
Best regards,<br>
Min<br>
<br>
<div class="m_-9136802374877624834moz-cite-prefix">On 2018/11/04 10:27, Zhifeng Yang via discuss wrote:<br>
</div>
<blockquote type="cite">
<div dir="ltr">
<div dir="ltr">Hi
<div><br>
</div>
<div>After I installed mpich and used it in a FORTRAN code. There is an error while running this FORTRAN code called real.exe by using the following command</div>
<div><br>
</div>
<div>$./real.exe</div>
<div>or</div>
<div>$mpirun ./real.exe</div>
<div>real.exe: error: _get_addr: No such file or directory</div>
<div>real.exe: error: _get_addr: No such file or directory</div>
<div>real.exe: error: _get_addr: No such file or directory</div>
<div>Fatal error in MPI_Init: Other MPI error, error stack:</div>
<div>MPIR_Init_thread(784).....:</div>
<div>MPID_Init(1323)...........: channel initialization failed</div>
<div>MPIDI_CH3_Init(120).......:</div>
<div>MPID_nem_init_ckpt(852)...:</div>
<div>MPIDI_CH3I_Seg_commit(364): PMI_Barrier returned -1</div>
<div> </div>
<div>But while using mpirun with specifying number of processors. as follows</div>
<div>$mpirun -np 1 ./real.exe</div>
<div>There is no error any more. I am not sure why? do you have any explanation? Thank you very much</div>
<div><br>
</div>
<div>Best regards</div>
<div>Zhifeng</div>
<div><br>
</div>
</div>
</div>
<br>
<fieldset class="m_-9136802374877624834mimeAttachmentHeader"></fieldset> <br>
<pre>_______________________________________________
discuss mailing list     <a class="m_-9136802374877624834moz-txt-link-abbreviated" href="mailto:discuss@mpich.org" target="_blank">discuss@mpich.org</a>
To manage subscription options or unsubscribe:
<a class="m_-9136802374877624834moz-txt-link-freetext" href="https://lists.mpich.org/mailman/listinfo/discuss" target="_blank">https://lists.mpich.org/mailman/listinfo/discuss</a>
</pre>
</blockquote>
<br>
</div>
_______________________________________________<br>
discuss mailing list     <a href="mailto:discuss@mpich.org" target="_blank">discuss@mpich.org</a><br>
To manage subscription options or unsubscribe:<br>
<a href="https://lists.mpich.org/mailman/listinfo/discuss" target="_blank">https://lists.mpich.org/mailman/listinfo/discuss</a><br>
</div>
</blockquote>
</div>
<br>
</div>
</div>
_______________________________________________<br>
discuss mailing list     <a href="mailto:discuss@mpich.org" target="_blank">discuss@mpich.org</a><br>
To manage subscription options or unsubscribe:<br>
<a href="https://lists.mpich.org/mailman/listinfo/discuss" target="_blank">https://lists.mpich.org/mailman/listinfo/discuss</a><br>
</div>
</blockquote>
</div>
<br>
</div>
</div>
</div>
_______________________________________________<br>
discuss mailing list     <a href="mailto:discuss@mpich.org" target="_blank">discuss@mpich.org</a><br>
To manage subscription options or unsubscribe:<br>
<a href="https://lists.mpich.org/mailman/listinfo/discuss" target="_blank">https://lists.mpich.org/mailman/listinfo/discuss</a><br>
</div>
</blockquote>
</div>
<br>
</div>
</div>
</div>
</blockquote>
</div>
<br>
</div>
</div>

_______________________________________________<br>
discuss mailing list     <a href="mailto:discuss@mpich.org" target="_blank">discuss@mpich.org</a><br>
To manage subscription options or unsubscribe:<br>
<a href="https://lists.mpich.org/mailman/listinfo/discuss" rel="noreferrer" target="_blank">https://lists.mpich.org/mailman/listinfo/discuss</a><br>
</blockquote></div>