Differences

This shows you the differences between two versions of the page.

--- doku:forge [2020/09/24 12:12] – sh
+++ doku:forge [2022/11/04 10:14] (current) – [Usage of map:] goldenberg
@@ Line 1: / Line 1: @@
-===== forge = map + ddt (VSC-3) =====
+===== forge = map + ddt =====
 ==== Synopsis: ====
 <html> <font color=#cc3300><b>map</b></font> </html> and <html> <font color=#cc3300><b>ddt</b></font> </html> are ARM's (formerly Allinea's)  advanced tools for performance analysis, see [[https://developer.arm.com/tools-and-software/server-and-hpc/debug-and-profile/arm-forge]].
-Licenses for up to 512 parallel tasks are available. Of additional note, <html> <font color=#cc3300><b>perf-report</b></font> </html> --- a related lightweight profiling tool --- has now been integrated into forge in more recent releases.
+Licenses for up to 512 parallel tasks are available. Of additional note, [[doku:perf-report|perf-report]] --- a related lightweight profiling tool --- has now been integrated into forge in more recent releases.
@@ Line 13: / Line 13: @@
    #SBATCH -J map
    #SBATCH -N 4
-   #SBATCH -L allinea@vsc
    #SBATCH --ntasks-per-node 16
    #SBATCH --ntasks-per-core  1
    module purge
-   module load  intel/18  intel-mpi/2018 allinea/20.1_FORGE
+   module load  intel/18  intel-mpi/2018 arm/20.1_FORGE
    map --profile srun --jobid $SLURM_JOB_ID --mpi=pmi2 -n 64 ./a.out
@@ Line 24: / Line 23: @@
 which generates a *.map file (note the mention of #tasks and #nodes together with the date/time stamp in the filename) that may then be analyzed via the gui, ie
-   ssh vsc3.vsc.ac.at -l my_uid -X
+   ssh vsc4.vsc.ac.at -l my_uid -X
    cd wherever/the/map/file/may/be
@@ Line 70: / Line 69: @@
    ssh vsc3.vsc.ac.at -l my_uid -X
    my_uid@l33$  cd wherever/my/app/may/be
-   my_uid@l33$  mpiicc -g -O0 my_app.c
    my_uid@l33$  salloc -N 4 -L allinea@vsc
-   my_uid@l33$  srun --jobid $SLURM_JOB_ID -n 4 hostname | tee ./machines.txt ( this is important ! it looks like a redundant command but will actually fix a lot of the prerequisites usually taken care of in the SLURM prologue of regular submit scripts, one of them being provisioning of required licenses )
+   my_uid@l33$  echo $SLURM_JOB_ID    ( just to figure out the current job ID, say it's 8909346 )
+   my_uid@l33$  srun --jobid 8909346 -n 4 hostname | tee ./machines.txt ( this is important ! it looks like a redundant command but will actually fix a lot of the prerequisites usually taken care of in the SLURM prologue of regular submit scripts, one of them being provisioning of required licenses )
                 ... let's assume we got n305-[044,057,073,074] which should now be listed inside file 'machines.txt'
    my_uid@l33$  rm -rf ~/.allinea/   ( to get rid of obsolete configurations from previous sessions )
    my_uid@l33$  module purge
    my_uid@l33$  module load  intel/18  intel-mpi/2018  allinea/20.1_FORGE   ( or whatever else suite of MPI )
+   my_uid@l33$  mpiicc -g -O0 my_app.c
    my_uid@l33$  ddt &     ( gui should open )
                 ... select 'Remote Launch - Configure'
@@ Line 85: / Line 85: @@
                 ... click OK twice to close the dialogues
                 ... click Close to exit from the Configure menu
-                ... next really select 'Remote Launch' by clicking the name tag that was auto-assigned above   ( licence label should be ok in the lower left corner and the connecting client should appear in the lower right corner )
+                ... next really select 'Remote Launch' by clicking the name tag that was auto-assigned above   ( licence label should be ok in the lower left corner and the hostname of the connecting client should appear in the lower right corner )
    ssh vsc3.vsc.ac.at -l my_uid   ( a second terminal will be needed to actually start the debug session )
-   my_uid@l34$  ssh n12-045       ( log into that compute node that was selected/prepared above for remote launch )
+   my_uid@l34$  ssh n305-044       ( log into that compute node that was selected/prepared above for remote launch )
-   my_uid@n12-045$  module load  intel/16  intel-mpi/5  allinea/18.2_FORGE
+   my_uid@n305-044$  module purge
-   my_uid@n12-045$  cd wherever/my/app/may/be
+   my_uid@n305-044$  module load  intel/18  intel-mpi/2018  allinea/20.1_FORGE
-   my_uid@n12-045$  mpirun -np 16  -machinefile ./machines.txt  hostname   ( just a dummy check to see whether all is set up correctly )
+   my_uid@n305-044$  cd wherever/my/app/may/be
-   my_uid@n12-045$  ddt --connect mpirun -np 64  -machinefile ./machines.txt ./a.out -arg1 -arg2   ( in the initial ddt-window a dialogue will pop up prompting for a Reverse Connection request; accept it and click Run and the usual debug session will start )
+   my_uid@n305-044$  srun --jobid 8909346 -n 16 hostname    ( just a dummy check to see whether all is set up and working correctly )
+   my_uid@n305-044$  ddt --connect srun --jobid 8909346 --mpi=pmi2 -n 64 ./a.out -arg1 -arg2   ( in the initial ddt-window a dialogue will pop up prompting for a Reverse Connection request; accept it and click Run and the usual debug session will start )
 ==== Further Reading: ====
-''/opt/sw/x86_64/glibc-2.17/ivybridge-ep/allinea/18.2_FORGE/doc/userguide-forge.pdf''
+''/opt/sw/x86_64/glibc-2.17/ivybridge-ep/allinea/20.1_FORGE/doc/userguide-forge.pdf''