Hello, I have been using Htcondor with MPICH 2 but I update MPICH 2 to MPICH 3 and now the MPD process manager has been deprecated and Hydra is de default process manager. The script I was using use some commands like mpdtrace, mpdallexit but they are not supported anymore. Here is my script, is it any possibility to change this script code to be able to work with Hydra? Thank you very much, Livan Valladares Martell Script: #!/bin/sh ##************************************************************** ## ## Copyright (C) 1990-2014, Condor Team, Computer Sciences Department, ## University of Wisconsin-Madison, WI. ## ## Licensed under the Apache License, Version 2.0 (the "License"); you ## may not use this file except in compliance with the License. You may ## obtain a copy of the License at ## ## http://www.apache.org/licenses/LICENSE-2.0 ## ## Unless required by applicable law or agreed to in writing, software ## distributed under the License is distributed on an "AS IS" BASIS, ## WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. ## See the License for the specific language governing permissions and ## limitations under the License. ## ##************************************************************** # Set this to the bin directory of MPICH installation MPDIR=/opt/local/bin PATH=$MPDIR:.:$PATH export PATH _CONDOR_PROCNO=$_CONDOR_PROCNO _CONDOR_NPROCS=$_CONDOR_NPROCS # Remove the contact file, so if we are held and released # it can be recreated anew rm -f $CONDOR_CONTACT_FILE PATH=`condor_config_val libexec`/:$PATH # mpd needs a conf file, and it must be # permissions 0700 mkdir tmp MPD_CONF_FILE=`pwd`/tmp/mpd_conf_file export MPD_CONF_FILE ulimit -c 0 # If you have a shared file system, maybe you # want to put the mpd.conf file in your home # directory echo "password=somepassword" > $MPD_CONF_FILE chmod 0700 $MPD_CONF_FILE # If on the head node, start mpd, get the port and host, # and condor_chirp it back into the ClassAd # so the non-head nodes can find the head node. if [ $_CONDOR_PROCNO -eq 0 ] then mpd > mpd.out.$_CONDOR_PROCNO 2>&1 & sleep 1 host=`mpdtrace -l | sed 1q | tr '_' ' ' | awk '{print $1}'` port=`mpdtrace -l | sed 1q | tr '_' ' ' | awk '{print $2}'` condor_chirp set_job_attr MPICH_PORT $port condor_chirp set_job_attr MPICH_HOST \"$host\" num_hosts=1 retries=0 while [ $num_hosts -ne $_CONDOR_NPROCS ] do num_hosts=`mpdtrace | wc -l` sleep 2 retries=`expr $retries + 1` if [ $retries -gt 100 ] then echo "Too many retries, could not start all $_CONDOR_NPROCS nodes, only started $num_hosts, giving up. Here are the hosts I could start " mpdtrace exit 1 fi done ## run the actual mpi job, which was the command line argument ## to the invocation of this shell script mpiexec -n $_CONDOR_NPROCS $@ e=$? mpdallexit sleep 20 echo $e else # If NOT the head node, acquire the host and port of # the head node retries=0 host=UNDEFINED while [ $host == "UNDEFINED" ] do host=`condor_chirp get_job_attr MPICH_HOST` sleep 2 retries=`expr $retries + 1` if [ $retries -gt 100 ]; then echo "Too many retries, could not get mpd host from condor_chirp, giving up." exit 1 fi done port=`condor_chirp get_job_attr MPICH_PORT` host=`echo $host | tr -d '"'` mpd --host=$host --port=$port > mpd.out.$_CONDOR_PROCNO 2>&1 fi _______________________________________________ discuss mailing list [email protected] To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss