Changes between Initial Version and Version 1 of aurora


Ignore:
Timestamp:
01/09/19 16:08:25 (7 years ago)
Author:
schlegel
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • aurora

    v1 v1  
     1== Getting an account ==
     2
     3more to come...
     4== ssh configuration ==
     5
     6You can add the following lines to `~/.ssh/config` on your local machine:
     7{{{
     8#!sh
     9Host aurora aurora.jpl.nasa.gov
     10   HostName aurora.jpl.nasa.gov
     11  User YOURUSERNAME
     12  HostKeyAlias aurora.jpl.nasa.gov
     13  HostbasedAuthentication no
     14}}}
     15and replace `YOURUSERNAME` by your JPL username.
     16
     17Once this is done, you can ssh aurora by simply doing:
     18
     19{{{
     20#!sh
     21ssh aurora
     22}}}
     23
     24== Password-less ssh ==
     25
     26Once you have the account, you can setup a public key authentication in order to avoid having to input your password for each run.
     27You need to have a SSH public/private key pair. If you do not, you can create a SSH public/private key pair by typing the following command and following the prompts (no passphrase necessary):
     28{{{
     29#!sh
     30$your_localhost% ssh-keygen -t rsa
     31Generating public/private rsa key pair.
     32Enter file in which to save the key (/Users/username/.ssh/id_rsa):RETURN
     33Enter passphrase (empty for no passphrase):RETURN
     34Enter same passphrase again:RETURN
     35Your identification has been saved in /Users/username/.ssh/id_rsa.
     36Your public key has been saved in /Users/username/.ssh/id_rsa.pub.
     37}}}
     38
     39Two files were created: your private key `/Users/username/.ssh/id_rsa`, and the public key `/Users/username/.ssh/id_rsa.pub`. The private key is read-only and only for you, it is used to decrypt all correspondence encrypted with the public key. The contents of the public key need to be copied to `~/.ssh/authorized_keys` on your aurora or halo account:
     40
     41{{{
     42#!sh
     43$your_localhost%scp ~/.ssh/id_rsa.pub username@your_remosthost:~
     44}}}
     45
     46Now on '''aurora''', copy the content of id_rsa.pub:
     47
     48{{{
     49#!sh
     50$your_remosthost%cat ~/id_rsa.pub >> ~/.ssh/authorized_keys
     51$your_remosthost%rm ~/id_rsa.pub
     52}}}
     53
     54== Environment ==
     55
     56On aurora, add the following lines to `~/.bash_login`:
     57{{{
     58#!sh
     59export ISSM_DIR=PATHTOTRUNK
     60source $ISSM_DIR/etc/environment.sh
     61source /usr/share/Modules/init/bash
     62module load intel/cluster-toolkit-2013.5.192
     63module load apps/matlab-2016b
     64
     65export PATH="$PATH:~/bin:$ISSM_DIR/scripts"
     66export LD_LIBRARY_PATH="$LD_LIBRARY_PATH:/opt/intel/impi/4.1.3/intel64/lib/"
     67}}}
     68
     69''Log out and log back in'' to apply this change.
     70
     71== Installing ISSM on aurora ==
     72
     73aurora can be used to run the code or with MATLAB locally. You can check out ISSM and install the following packages:
     74 - autotools
     75 - PETSc (use the aurora script and ``follow`` the instructions, you will need to submit a job and compile PETSc manually, do not make test, it will not work on the cluster)
     76 - m1qn3
     77
     78Use the following configuration script (adapt to your needs):
     79
     80{{{
     81#!sh
     82./configure \
     83   --prefix=$ISSM_DIR \
     84   --with-matlab-dir=$MATLAB_DIR \
     85   --with-m1qn3-dir=$ISSM_DIR/externalpackages/m1qn3/install \
     86   --with-metis-dir=$ISSM_DIR/externalpackages/petsc/install \
     87   --with-petsc-dir=$ISSM_DIR/externalpackages/petsc/install \
     88   --with-scalapack-dir=$ISSM_DIR/externalpackages/petsc/install \
     89   --with-mpi-include="/opt/intel/impi/4.1.3/intel64/include/" \
     90   --with-mpi-libflags="-L/opt/intel/impi/4.1.3/intel64/lib/ -lmpi -lmpiif" \
     91   --with-petsc-arch=$ISSM_ARCH \
     92   --with-mkl-libflags="-L/opt/intel/composer_xe_2013.5.192/mkl/lib/intel64/ -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread -lm" \
     93   --with-mumps-dir=$ISSM_DIR/externalpackages/petsc/install \
     94   --with-fortran-lib="-L/opt/intel/composer_xe_2013.5.192/compiler/lib/intel64/ -lifcore -lifport" \
     95   --with-triangle-dir=$ISSM_DIR/externalpackages/triangle/install \
     96   --with-cxxoptflags="-O3" \
     97   --with-vendor=intel-aurora \
     98   --enable-development \
     99   --enable-debugging
     100}}}
     101
     102== aurora_settings.m ==
     103
     104To launch from your local computer, you have to add a file in `$ISSM_DIR/src/m` entitled `aurora_settings.m` with your personal settings on your local issm install:
     105
     106{{{
     107#!m
     108cluster.login='schlegel';
     109cluster.codepath='/home/schlegel/issm/trunk/bin';
     110cluster.executionpath='/home/schlegel/issm/trunk/execution';
     111}}}
     112
     113Use your username for the `login` and enter your code path and execution path. These settings will be picked up automatically by matlab when you do `md.cluster=aurora()`
     114
     115Note that the `executionpath' creates temporary binary files that can be removed once the job is complete. For this reason, you can set the path to be somewhere on the /aurora_nobackup/issm or /halo_nobackup/issm filesystem, which is unlimited temporary storage on these systems.
     116
     117
     118== Running jobs on aurora  ==
     119
     120On aurora, you can use up to 256 cpus. The more nodes and the longer the requested time, the more you will have to wait in the queue. So choose your settings wisely:
     121[https://hpcs.jpl.nasa.gov/Users/HPC_resources.html]
     122
     123If you are running from your local machine:
     124 {{{
     125#!m
     126md.cluster=aurora();
     127}}}
     128
     129This will default to 1 node, 24 cpus.
     130
     131
     132Before you run your job, make sure to open a port first and enter the port number in md.cluster. Here is a handy alias:
     133
     134{{{
     135#!sh
     136alias auroratunnel='ssh -L 1099:localhost:22 aurora'
     137}}}
     138That will open port number 1099 that you can then use in ISSM so that you don't need to enter your password.
     139
     140
     141Or, if you are launching from MATLAB on aurora/halo:
     142 {{{
     143#!m
     144md.cluster=localaurora();
     145}}}
     146
     147
     148To submit a job on aurora, do:
     149
     150 {{{
     151#!m
     152qsub job.queue
     153}}}
     154
     155Now if you want to check the status of your job and the queue you are using, type in the bash with the aurora/halo session:
     156
     157 {{{
     158#!sh
     159qstat-u USERNAME
     160}}}
     161
     162You can delete your job manually by typing:
     163
     164{{{
     165#!sh
     166qdel JOBID
     167}}}
     168
     169where 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.