This session will cover following topics

...

 1. Connecting to remote working space

1.1 Connecting to ITER cluster

...

You will probably want to use an SSH client that supports remote X for graphical environment. The host hpc-login4.iter.org supports NX clients such as NoMachine whilst you can connect to hpc-app1.iter.org with X2go.
You can use your ITER username and password, which is the same for the collaborative websites

1.2 Connection to the EUROfusion Gateway

In order to use Gateway, you have to meet following requirements:

• You have to have Gateway account (in case you don't have one, you have to request it) - refer to this link
• You have to install NX client

...

...

...

2.Setting IMAS

Loading all the IMAS modules.

...

You will get access to libraries and most commonly used tools (FC2K, Kepler, Matlab interfaces, Python interfaces, etc.)

2.1

...

Before you begin, make sure to backup all the Kepler related data in your $HOME directory

shell> cd $HOME
shell> mv kepler kepler_09.2016
shell> mv .kepler .kepler_09.2016
shell> mv KeplerData KeplerData_09.2016

After you have backed up your data, install Kepler

2.2. Kepler installation

In order to install most recent release of Kepler you should execute following commands:

• Set-up KEPLER_DIR system variable

  Code Block
  shell> mkdir -p $ITMWORK/mykeplers shell> setenv KEPLER_DIR $ITMWORK/mykeplers

• List all available Kepler releases

  Code Block
  shell> list_keplers.sh -all List of available Kepler installations Local installations - local (default) - R3.9/local - R3.10/local - ETS4JET-training/local Central installations - central (default) - rc (release candidate) - R3.9/central - R3.10/central - ETS4JET-training/central

• Install chosen version of Kepler

  Code Block
  shell> install_kepler.sh R3.10_central R3.10/central 'tutorial Kepler installation'

  Info
  title install_kepler.sh
  shell> install_kepler.sh [target dir] [source Kepler] [description] * target dir : chosen directory inside KEPLER_DIR * source Kepler : name of Kepler release (list available through 'list_kepler.sh -all') - local : default Kepler (local) - central : default Kepler (central) - rc : release candidate Kepler (central) - R#.#/local : tagged Kepler version R#.# (local) - R#.#/central : tagged Kepler version R#.# (central) * description : chosen description (e.g. 'this is my test kepler')

• List all your Kepler installations

  Code Block
  shell> list_my_keplers.sh Using location at: /pfs/work/bpalak/my_keplers Following kepler installations are ready for use selected type name - 'description' [ ] [Local ] kepler_3.10_loc - 'r3.10 local' [ ] [Central ] kepler_ets - '' [ ] [Central ] R3.10_central - 'tutorial Kepler installation' [x] [Central ] R3_RC_ETS_central - ''

• Switch to chosen Kepler

  Code Block
  bash> switch_to_kepler.sh R3.10_central

3. Setting up tutorial environment

. Setting IMAS @ ITER Cluster

The fastest way to setup the environment for Kepler, IMAS and all the other dependencies is to load kepler module

module load fc2k kepler 

 Please follow this  >link<  to get detailed instructions how to set-up IMAS environment on ITER cluster

2.2. Setting IMAS @ EUROfusion Gateway

You can switch to IMAS environment using:

module load imasenv

2.3. Setting IMAS @ JET (Freia)

You can switch to IMAS environment using:

module load imas-modules    # add all imas modules to module path
module load imas/3.17.1     # or other version of your choice; default is older
module swap mdsplus/7.7.11  # the default version 6.1 loaded by JAC module is too slow with IMAS
module load pyual

The JET installation is maintained by Jonathan Hollocombe.

3. Database initialization

...

In order to work with MDSPlus database you have to create and initialize MDSPlus database structure (details can be found here). This can be done by issuing following command:

Code Block
$ITMSCRIPTDIR/create_user_itm_dir test 4.10b

After this command is executed, you will be able to locate MDSPlus files under following location: $HOME/public/itmdb/. This location is a link to your $ITMWORK file system, where actual data can be found.

At this point you have created your database structure inside $ITMWORK file system. This is required for further parts of the tutorial.

You should execute an ITM initialization script:

Code Block
source $ITMSCRIPTDIR/ITMv1 kepler test 4.10b > /dev/null

Info
title Adding source script into .login
In order to make sure that source script is always started when you log in, add it to ~/.login echo "\n" >> ~/.login echo "source $ITMSCRIPTDIR/ITMv1 kepler test 4.10b > /dev/null" >> ~/.login

 

After this script is executed you will have an access to following tools:

• FC2K
• HPC2K
• Kepler
• jTraverser
• ISE

These tools will be discussed later in the tutorial.

imasdb <machine_name> e.g.: imasdb test

4. Getting tutorial sources and input data

4.1. Copying examples

Copy tutorial examples. You can install them by issuing following command from Terminal.

4.1.

...

1 IMAS @ ITER

4.1.2 IMAS @ EUROfusion Gateway

...

...

...

4.

...

2. Copying input data

Copy data files used throughout the tutorial:

...

4.2.1 IMAS @ ITER

4.2.2 IMAS @ EUROfusion Gateway

...

...

...

...

...

4.

...

3. Set TUTORIAL_DIR variable

...

4.3.1 IMAS @ ITER

shell> setenvexport TUTORIAL_DIR $ITMWORK=$HOME/tutorials/20162018-09-Ljubljana

5. Varia

JVMJ9VM003W JIT compiler "/afs/rzg/@sys/soft/java/jdk6/bin/javac" not found. Will use interpreter. 

...

03-ITER

4.3.2 IMAS @ EUROfusion Gateway

shell> setenv TUTORIAL_DIR $HOME/tutorials/2018-03-ITER

5. Java memory settings

Please do not forget to set JAVA memory settings:

export _JAVA_OPTIONS="-Xss20m -Xms1g -Xmx4g"