[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[HTCondor-users] Support Required


1. We are running our HTCondor grid on 4 machines as a test case. Each machine has 4 cores. total number of cores in the grid are 16. 3 machines are rack servers with xeon processor each having 4 cores (i.e total 12 cores) and one machine is Intel core i5 workstation having 4 cores also. When we run job on 3 servers, everything is fine and executes properly. But when we include workstation (Intel core i5 ) in the grid and execute the job, the job remains idle and does not execute.

2. When we execute an MPI program as mentioned above on three racks servers. It shows in log file that it executed on all 3 machines but it doesn't utilizes the processing power of all 3 machines equally instead it used processing power of 1 machine.

 Kindly help me in this regard. (Submit file is mentioned below)

universe = parallel
executable = mpi_prime.exe
Error = err.$(Cluster).$(Process)
Output = out.$(Cluster).$(Process)
Log = log.$(Cluster).$(Process)
should_transfer_files = yes
when_to_transfer_output = on_exit
+ParallelShutdownPolicy = "WAIT_FOR_ALL"
machine_count = 16