<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html;
      charset=windows-1252">
  </head>
  <body>
    <p>One should keep in mind that sacct results for memory usage are
      not accurate for Out Of Memory (OoM) jobs.  This is due to the
      fact that the job is typically terminated prior to next sacct
      polling period, and also terminated prior to it reaching full
      memory allocation.  Thus I wouldn't trust any of the results with
      regards to memory usage if the job is terminated by OoM.  sacct
      just can't pick up a sudden memory spike like that and even if it
      did  it would not correctly record the peak memory because the job
      was terminated prior to that point.</p>
    <p><br>
    </p>
    <p>-Paul Edmon-</p>
    <p><br>
    </p>
    <div class="moz-cite-prefix">On 3/15/2021 1:52 PM, Chin,David wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:BN7PR01MB3858E1245783056E36FC013CD66C9@BN7PR01MB3858.prod.exchangelabs.com">
      <meta http-equiv="Content-Type" content="text/html;
        charset=windows-1252">
      <style type="text/css" style="display:none;">P {margin-top:0;margin-bottom:0;}</style>
      <div style="font-family: "Courier New", monospace;
        font-size: 12pt; color: rgb(0, 0, 0);">
        Hi, all:</div>
      <div style="font-family: "Courier New", monospace;
        font-size: 12pt; color: rgb(0, 0, 0);">
        <br>
      </div>
      <div style="font-family: "Courier New", monospace;
        font-size: 12pt; color: rgb(0, 0, 0);">
        I'm trying to understand why a job exited with an error
        condition. I think it was actually terminated by Slurm: job was
        a Matlab script, and its output was incomplete. </div>
      <div style="font-family: "Courier New", monospace;
        font-size: 12pt; color: rgb(0, 0, 0);">
        <br>
      </div>
      <div style="font-family: "Courier New", monospace;
        font-size: 12pt; color: rgb(0, 0, 0);">
        Here's sacct output:</div>
      <div style="font-family: "Courier New", monospace;
        font-size: 12pt; color: rgb(0, 0, 0);">
        <br>
      </div>
      <div style="font-family: "Courier New", monospace;
        font-size: 12pt; color: rgb(0, 0, 0);">
                       JobID    JobName      User  Partition      
         NodeList    Elapsed      State ExitCode     ReqMem     MaxRSS
         MaxVMSize                        AllocTRES AllocGRE
        <div>-------------------- ---------- --------- ----------
          --------------- ---------- ---------- -------- ----------
          ---------- ---------- --------------------------------
          --------</div>
        <div>               83387 ProdEmisI+      foob        def      
            node001   03:34:26 OUT_OF_ME+    0:125      128Gn          
                              billing=16,cpu=16,node=1</div>
        <div>         83387.batch      batch                            
           node001   03:34:26 OUT_OF_ME+    0:125      128Gn   1617705K
            7880672K              cpu=16,mem=0,node=1</div>
                83387.extern     extern                            
         node001   03:34:26  COMPLETED      0:0      128Gn       460K  
         153196K         billing=16,cpu=16,node=1<br>
      </div>
      <div style="font-family: "Courier New", monospace;
        font-size: 12pt; color: rgb(0, 0, 0);">
        <br>
      </div>
      <div>
        <div style="font-family: "Courier New", monospace;
          font-size: 12pt; color: rgb(0, 0, 0);">
          Thanks in advance,</div>
        <div style="font-family: "Courier New", monospace;
          font-size: 12pt; color: rgb(0, 0, 0);">
              Dave</div>
        <div style="font-family: "Courier New", monospace;
          font-size: 12pt; color: rgb(0, 0, 0);">
          <br>
        </div>
        <div id="Signature">
          <div>
            <div id="divtagdefaultwrapper" dir="ltr"
              style="font-size:12pt; color:#000000; font-family:'Courier
              New',monospace">
              <div class="BodyFragment"><font size="2"><span
                    style="font-size:10pt">
                    <div class="PlainText"
                      style="font-family:"Courier
                      New",monospace; font-size:13.3333px">
                    </div>
                    <span id="ms-rterangepaste-start"></span>
                    <div>--</div>
                    <div>
                      <div>David Chin, PhD (he/him)   Sr. SysAdmin,
                        URCF, Drexel</div>
                      <div><a class="moz-txt-link-abbreviated" href="mailto:dwc62@drexel.edu">dwc62@drexel.edu</a>                   
                         215.571.4335 (o)</div>
                      <div>For URCF support: <a class="moz-txt-link-abbreviated" href="mailto:urcf-support@drexel.edu">urcf-support@drexel.edu</a></div>
                      <div><a class="moz-txt-link-freetext" href="https://proteusmaster.urcf.drexel.edu/urcfwiki">https://proteusmaster.urcf.drexel.edu/urcfwiki</a></div>
                      <div>github:prehensilecode</div>
                    </div>
                    <span id="ms-rterangepaste-end"></span>
                    <div class="PlainText"><br>
                    </div>
                  </span></font></div>
            </div>
          </div>
        </div>
      </div>
      <br>
      <p
        style="font-family:Calibri;font-size:10pt;color:#000000;margin:5pt;"
        align="Left">
        Drexel Internal Data<br>
      </p>
    </blockquote>
  </body>
</html>