Running and Verification

Procedure

1. Use PuTTY to log in to the server as the root user.
2. Run the following command to decompress the case files:

   tar -xvf CMAQv5.3.1_Benchmark_2Day_Input.tar.gz -C /path/to/CMAQ/CMAQ_Project/data

3. Run the following command to switch to the test directory:

   cd /path/to/CMAQ/CMAQ_Project/CCTM/scripts

4. Configure the running script.
   1. Open the running script.

      vim run_cctm_Bench_2016_12SE1.csh

   2. Press i to enter the insert mode.
      Change the value of compiler to gcc and the value of compilerVrsn to 9.3.0:

      setenv compiler gcc
      setenv compilerVrsn 9.3.0

      Change the values of NPCOL and NPROW. The product of the two values is equal to the number of parallel cores. For example:

      @ NPCOL  =  12; @ NPROW =  8

      Modify the INPDIR path, for example:

      setenv INPDIR /path/to/CMAQ/CMAQ_Project/data/CMAQv5.3.1_Benchmark_2Day_Input/2016_12SE1

      Modify the Open MPI running parameters. For example:

      /usr/bin/time -p mpirun -np $NPROCS -x LD_PRELOAD=/path/to/MATH/lib/libmathlib.so --allow-run-as-root --mca btl ^openib $BLD/$EXEC

   3. Press Esc, type :wq!, and press Enter to save the file and exit.
5. Run the following command to perform the test:

   csh -x /path/to/CMAQ/CMAQ_Project/CCTM/scripts/run_cctm_Bench_2016_12SE1.csh 2>&1 | tee -a /path/to/CMAQ/CMAQ_Project/CCTM/scripts/csh.log

   Check the value of Total Time in the csh.log file. The unit is second. A smaller value indicates better performance.

   Figure 1 shows an example of the running result.

   Figure 1 Example result