https://community.intel.com/t5/Intel-MPI-Library/mpicleanup-in-MPI-4-0-3/m-p/774985#M270 Hi Mateo,<BR /><BR />It generally takes a pretty bad failure for cleanup to not automatically occur. If you want to see the file, I would recommend starting a job that is either long enough or needs user input. While the job is waiting, you can look at the file and try mpicleanup. By default, the file should be in /tmp, in the form mpiexec_${username}_$PPID.log and will contain the PIDs of each process grouped by host.<BR /><BR />Sincerely,<BR />James Tullos<BR />Technical Consulting Engineer<BR />Intel Cluster Tools Thu, 02 Aug 2012 19:24:22 GMT James_T_Intel 2012-08-02T19:24:22Z https://community.intel.com/t5/Intel-MPI-Library/mpicleanup-in-MPI-4-0-3/m-p/774984#M269 Hi,<DIV></DIV><DIV>I am trying MPI 4.0.3 and would like to know which is the best way to cleanup the environment after an mpirun execution using Hydra finishes incorreclty.</DIV><DIV></DIV><DIV></DIV><DIV>In previous versions (with MPDs) I was doing mpdallexit to ensure all daemons where down.</DIV><DIV></DIV><DIV></DIV><DIV>I tried usingI_MPI_HYDRA_CLEANUP environment variable to generate the input file for the mpicleanup tool, but it seems this file is never created.</DIV><DIV></DIV><DIV></DIV><DIV>This is how I am running mpi:</DIV><DIV></DIV><DIV></DIV><DIV>mpirun -r ssh -f /home/mguzman/wdir/mpd.host -genv I_MPI_FABRICS shm:dapl -env I_MPI_HYDRA_CLEANUP=1 -machinefile /home/mguzman/wdir/machinefile -n 16 /home/mguzman/wdir/test.impi</DIV><DIV></DIV><DIV></DIV><DIV>I have read that this file is erased when the job ends, so I tryed to look for that file while the command is being executed, but culdn't find it.</DIV><DIV></DIV><DIV> <DIV id="_mcePaste">Am I doing something wrong? How can I force mpi to fail in order to try mpicleanup tool?</DIV></DIV><DIV id="_mcePaste"></DIV><DIV id="_mcePaste"></DIV><DIV id="_mcePaste">Thanks in advance</DIV><DIV id="_mcePaste"></DIV><DIV id="_mcePaste"></DIV><DIV id="_mcePaste">Mateo Guzman</DIV><DIV></DIV><DIV></DIV> Thu, 02 Aug 2012 15:46:36 GMT https://community.intel.com/t5/Intel-MPI-Library/mpicleanup-in-MPI-4-0-3/m-p/774984#M269 mateoguzman 2012-08-02T15:46:36Z https://community.intel.com/t5/Intel-MPI-Library/mpicleanup-in-MPI-4-0-3/m-p/774985#M270 Hi Mateo,<BR /><BR />It generally takes a pretty bad failure for cleanup to not automatically occur. If you want to see the file, I would recommend starting a job that is either long enough or needs user input. While the job is waiting, you can look at the file and try mpicleanup. By default, the file should be in /tmp, in the form mpiexec_${username}_$PPID.log and will contain the PIDs of each process grouped by host.<BR /><BR />Sincerely,<BR />James Tullos<BR />Technical Consulting Engineer<BR />Intel Cluster Tools Thu, 02 Aug 2012 19:24:22 GMT https://community.intel.com/t5/Intel-MPI-Library/mpicleanup-in-MPI-4-0-3/m-p/774985#M270 James_T_Intel 2012-08-02T19:24:22Z