Note that this information is up to date as of ISSM trunk-jpl version 24100

== Getting an account ==

New users have to ask the PIs (Eric Larour or Mathieu Morlighem) to request a NASA account for them if they don't have one already:
 - go to [https://www.nas.nasa.gov/hecc/portal/accounts] select option 3 ("I want to request a NASA identity for one of my new users")
 - provide the information requested

Then, a NCCS account has to be created by the new user:
 - go to [https://www.nccs.nasa.gov/nccs-users/instructional/account-set-up] and fill out a NAMS request
 - provide the information requested using either GID = s5692 for Eric's group or GID = s1507 for Mathieu's group
 - the PI will receive an email to approve the request
 - All users must complete NASA-mandatory Basic IT Security Training (This year's training is called "FY21 CYBERSECURITY AND SENSITIVE UNCLASSIFIED INFORMATION AWARENESS COURSE"). 
 - UCI IT Security officer: Josh Drummond jdrummon@uci.edu
 - JPL IT Security officer: Tomas Soderstrom Tomas.J.Soderstrom@jpl.nasa.gov

== Environment ==

Since you will be using `bash`, add the following to your `~/.bashrc`:
{{{
#ISSM
export ISSM_DIR="/home/jbondzio/ISSM/trunk-jpl"
source $ISSM_DIR/etc/environment.sh

export MATLAB_DIR="/discover/vis/mathworks/matlab_r2017b"
export MATLABBIN=$MATLAB_DIR"/bin"

#Packages
module purge
module load comp/intel/20.0.0.166
module load mpi/sgi-mpt/2.17
module load cmake/3.17.0
module load cdo/1.9.9rc2
module load matlab/R2017b
}}}

And replace `ISSM_DIR` with your actual trunk. ''Log out and log back in'' to apply this change.

== Installing ISSM on Discover ==

'''Do NOT install mpich'''. We have to use the one provided by NCCS. You can check out ISSM and install the following packages:
 - m1qn3
 - PETSc (use `install-3.14-discover.sh`)

You will need to run the following command before configuring ISSM:
{{{
#!sh
cd $ISSM_DIR
autoreconf -ivf
}}}

Use the following configuration script for ISSM (adapt to your needs):

{{{
#!sh
./configure \
 --prefix=$ISSM_DIR \
 --with-matlab-dir=$MATLAB_DIR \
 --with-m1qn3-dir=$ISSM_DIR/externalpackages/m1qn3/install \
 --with-triangle-dir=$ISSM_DIR/externalpackages/triangle/install \
 --with-metis-dir=$ISSM_DIR/externalpackages/petsc/install \
 --with-petsc-dir=$ISSM_DIR/externalpackages/petsc/install \
 --with-scalapack-dir=$ISSM_DIR/externalpackages/petsc/install \
 --with-mpi-include="/usr/local/sgi/mpi/mpt-2.17/include" \
 --with-mpi-libflags="-L/usr/local/sgi/mpi/mpt-2.17/lib/ -lmpi -lmpi++" \
 --with-petsc-arch=$ISSM_ARCH \
 --with-mkl-libflags="-L/usr/local/intel/2020/compilers_and_libraries_2020.0.166/linux/compiler/lib/intel64/ -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread -lm -limf -lsvml" \ 
 --with-mumps-dir=$ISSM_DIR/externalpackages/petsc/install \
 --with-fortran-lib="-L/usr/local/intel/2020/compilers_and_libraries_2020.0.166/linux/compiler/lib/intel64/ -lifcore" \
 --with-cxxoptflags="-O3" \
 --with-vendor=intel-discover \
 --enable-development \
 --enable-debugging
}}}

== Installing ISSM on Discover with Dakota ==

For Dakota to run, you you will still need to make PETSc, and m1qn3.

In addition, will need to build the external packages:
 - gsl, install-static.sh
 - boost, install-1.55-discover.sh
 - dakota, install-6.2-discover.sh
 - chaco, install.sh

Finally, you will need to add the following lines to your configuration script:

{{{
 --with-dakota-dir=$ISSM_DIR/externalpackages/dakota/install \
 --with-boost-dir=$ISSM_DIR/externalpackages/boost/install \
 --with-chaco-dir=$ISSM_DIR/externalpackages/chaco/install \
}}}

Remember, you will need to run the following commands before configuring ISSM:

{{{
cd $ISSM_DIR
autoreconf -ivf
}}}

== discover_settings.m ==

You have to add a file in `$ISSM_DIR/src/m` entitled `discover_settings.m` (on the machine you wish to access Discover) with your personal settings:

{{{
#!m
cluster.login='mmorligh';
cluster.queue='nccs';
cluster.codepath='/home/mmorligh/issm/trunk/bin';
cluster.executionpath='/discover/nobackup/mmorligh/execution/';
cluster.grouplist='s5692';
cluster.port=1099;
}}}

use your username for the `login` and enter your code path and execution path. Be sure to create the final execution directory (mkdir) within the nobackup folder. These settings will be picked up automatically by matlab when you do `md.cluster=discover()`. To determine your `grouplist`, on Discover run: 

{{{
%groups USERNAME
}}}

Make sure your module list includes comp/intel/20.0.0.166, mpi/sgi-mpt/2.17, and cmake/3.17.0

== Running jobs on Discover ==

On Discover, the more nodes and the longer the requested time, the more you will have to wait in the queue. So choose your settings wisely:

 {{{
#!m
md.cluster=discover('numnodes',8,'time',30,'processor','west','queue','debug');
md.cluster.time=10;
}}}

to have a maximum job time of 10 minutes and 8 westmere nodes. If the run lasts longer than 10 minutes, it will be killed and you will not be able to retrieve your results.

Now if you want to check the status of your job and the queue you are using, use this command:

 {{{
#!sh
qstat -u USERNAME
}}}

You can delete your job manually by typing:

{{{
#!sh
qdel JOBID
}}}

where JOBID is the ID of your job (indicated in the Matlab session). Matlab indicates too the directory of your job where you can find the files `JOBNAME.outlog` and `JOBNAME.errlog`. The outlog file contains the informations that would appear if you were running your job on your local machine and the errlog file contains the error information in case the job encounters an error. 

If you want to load results from the cluster manually (for example if you have an error due to an internet interruption), you find in the informations Matlab gave you `/home/srebuffi/trunk-jpl/execution//SOMETHING/JOBNAME.lock `, you copy the SOMETHING and you type in Matlab:

{{{
#!m
md=loadresultsfromcluster(md,'runtimename','SOMETHING');
}}}