Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

You may use the the ecinteractive tool to open interactive sessions with dedicated resources on the HPCF and ECS.

...

  • Only one interactive job is allowed at a time
  • Your job keeps on running after you exit the interactive shell, so you can reattach to it any time or open multiple interactive shells within the same job.
  • You may open a basic graphical desktop for X11 applications.
  • You may open a Jupyter Lab instance and connect to it through your browser., although use of JupyterHub is recommended instead.
  • If running on the HPCF platform, your interactive job may be scheduled to any of the four complexes. You may also choose to run on ECSBy default it will submit to the local cluster, or AA if run from the Linux VDI, although you can choose what complex (platform) to use.
  • You can run ecinteractive from any Atos HPCF complex, ECS, and Red Hat Linux VDI. You may also copy the script to your end user device and use it from there. It should work from Linux, Mac, or WSL under windows, and requires the Teleport tsh client to be installed and configured.
No Format
$ ecinteractive -h

Usage :  /usr/local/bin/ecinteractive [options] [--]

    -d|desktop     Submits a vnc job (default is interactive ssh job)
    -j|jupyter     Submits a jupyter job (default is interactive ssh job) 
    -J|jupyters     Submits a jupyter job with HTTPS support (default is interactive ssh job)

     More OptionsMore Options:
    -h|help        Display this message
    -v|version     Display script version
    -p|platform    Platform (default aahpc. Choices: aahpc, ab, ac, ad, ecs)
    -u|user        ECMWF User (default userusxa)
    -A|account     Project account
    -c|cpus        Number of CPUs (default 2)
    -m|memory      Requested Memory (default 8G8 GB)
    -s|tmpdirsize  Requested TMPDIR size (default 3 GB)
    -tg|timegpu         WallRequest clocka limitGPU (default 12:00:00limited availability)
    -kt|killtime        CancelWall anyclock runninglimit interactive job(default 12:00:00)
    -qf|queryforward     Ports  Check running jobto forward, comma separated (default NONE)
    -Qe|quietexport      Environment  Silent mode
    -o|outputvariables to export, comma separated (default NONE)
    -k|kill       Output fileCancel forany therunning interactive job (default /dev/null)
    -xq|query       Check running job
    -Q|quiet       Silent mode
   set -x -o|output      Output file for the interactive job (default /dev/null)
    -x             set -x


Note
titleBefore
Note
titleBefore you start: Set up your SSH key-based authentication

For ecinteractive to work properly, passwordless ssh must be configured between Atos HPCF nodes. See HPC2020: How to connect for more information on how to set it up.

...

You can get an interactive shell running on an allocated node within the Atos HCPF with just calling ecinteractive. By default it will just use the default settings which are:

...

Jobs will be submitted in either of these queues depending on if the user has access to the full HPC service (ni queue, for Member State users) or the ECS service (ei queue, for co-operating States, as service previously known as ECGATE) with the following default and maximum resource levels.

Multiexcerpt include
MultiExcerptNameecinteractive
PageWithExcerptUDOC:HPC2020: Batch system

...


If you need more resources, you may use the ecinteractive options when creating the job. For example, to get a shell with 4 cpus and 16 GB or memory for 12 hours:

...

No Format
[user@aa6-100 ~]$ ecinteractive -k
cancelling job 10225018...
CLUSTER JOBID       STATE       EXEC_HOST  TIME_LIMIT  TIME_LEFT MAX_CPUS MIN_MEMORY TRES_PER_NODE
aa      10225018    RUNNING     aa6-104      12:00:00   11:55:34        4        16G     ssdtmp:3G
Cancel job_id=10225018 name=user-ecinteractive partition=inter [y/n]? y
Connection to aa-login closed.y
Connection to aa-login closed.

Customising ecinteractive default resources

You may override the program defaults if you always create your ecinteractive sessions with specific resource requirements by defining them in ~/.ecinteractiverc. Here is a sample file with default values that may be used as a starting point:

Code Block
languagebash
title~/.ecinteractiverc
#Platform may be hpc or ecs
PLATFORM=hpc
# Number of CPUs
NCPUS=2
# Memory in GB
MEMORY=8
# TMPSIZE includes the SSDTMPDIR space
TMPSIZE=3
# Wall clock time limit for the session, in any valid SLURM time specification
TIME="12:00:00"

You don't need to define all the parameters, you may only define those you wish to change in respect to program defaults. 

Info
titleSettings precedence

Command line options always take precedence over default values, so you can always override those on specific invocations of ecinteractive with the required command line options.

Visual Studio Code integration

If you are a Visual Studio Code (VSCode) user, it is possible to use its Remote SSH extension to connect to your ecinteractive session instead of the login node to enjoy dedicated interactive resources.

Prerequisites

In order for this to work, you will need to have ecinteractive available on the platform where you run VSCode. If you are using your own computer, you need to follow these steps: 

  • Download the latest ecinteractive tool. Mac, and WSL are supported:
    No Format
    curl -so ecinteractive https://git.ecmwf.int/projects/USS/repos/ecinteractive/raw/ecinteractive
    chmod +x ecinteractive
    Once you have it, you can always make sure it is up to date running:
    No Format
    ./ecinteractive -U

  • Make sure Teleport is properly configured for you to be able to connect via ssh to our HPCF or ECS.

No installation is required if you are using VSCode and ecinteractive from the Linux VDI, as they are both available. 

Setup

You will need to make sure the following snippet is at the top of your ~/.ssh/config on the platform where you run VScode itself:

Code Block
languagebash
title~/.ssh/config
# ecinteractive dynamic host
Include ecinteractive_ssh_config

With that in place, every time you run ecinteractive from this platform, a  file ~/.ssh/ecinteractive_ssh_config will be created or updated, defining a host alias "ecinteractive" pointing to the actual host running your interactive session. 

Connecting VSCode to ecinteractive session

Whenever you want to use VSCode on your interactive session, you will need to run the ecinteractive command with the desired options from a separate terminal at least once, so that the "ecinteractive" host alias points to the right host. An example with default options to just update the host alias could be:

No Format
ecinteractive :

On VScode, a new host called ecinteractive will appear in your remote host list. If you select that one, a vscode remote session will be started using the resources from your ecinteractive session.

Info
titleActive sessions and up-to-date host alias

Remember that if your ecinteractive session expires or you kill it, you will need to rerun ecinteractive from the same platform where you run VSCode before you can use that connection again

Opening graphical applications within your interactive job

...

Excerpt

Opening a Jupyter Lab instance

Tip
titleJupyterHub Service

You may find it more convenient to use ECMWF's Jupyterhub instead of ecinteractive to run Jupyter on HPCF or ECS. Only ECMWF's Jupyterhub is the recommended way to run Jupyter at ECMWF, and only your browser is required to access this service.

You can use ecinteractive to open up a Jupyter Lab instance on the HPCF.  The application would effectively run on the allocated node for the job, and would allow you to conveniently interact with it from your browser. When running from VDI or your end user device, ecinteractive will try to open it in a new tab automatically. Alternatively you may manually open the URL provided to connect to your Jupyter Lab session.
No Format
[user@aa6-100 ~]$ ecinteractive -j
Using interactive job:
CLUSTER JOBID       STATE       EXEC_HOST  TIME_LIMIT  TIME_LEFT MAX_CPUS MIN_MEMORY TRES_PER_NODE
aa      10225277    RUNNING     aa6-104       6:00:00    5:58:07        2         8G     ssdtmp:3G

To cancel the job:         
        /usr/local/bin/ecinteractive -k

Attaching to Jupyterlab session...
To manually re-attach go to http://aa6-104.ecmwf.int:33698/?token=b1624da17308654986b1fd66ef82b9274401ea8982f3b747

To use your own conda environment as a kernel for Jupyter notebook you will need to have ipykernel installed in the conda environment before starting ecinteractive job. ipykernel can be installed with:

No Format
[user@aa6-100 ~]$ conda activate {myEnv}
[user@aa6-100 ~]$ conda install ipykernel
[user@aa6-100 ~]$ python3 -m ipykernel install --user --name={myEnv}

The same is true if you want to make your own Python virtual environment visible in Jupyterlab

No Format
[user@aa6-100 ~]$ source {myEnv}/bin/activate
[user@aa6-100 ~]$ pip3 install ipykernel
[user@aa6-100 ~]$ python3 -m ipykernel install --user --name={myEnv}

To remove your personal kernels from Jupyterlab once you don't need them anymore, you could do so with:

No Format
jupyter kernelspec uninstall {myEnv}


Expand
titleHow to connect to the Jupiter Lab instance from web browser on your end user device
  • Authenticate with Teleport using jump.ecmwf.int as described in the  Teleport SSH Access page
  • Copy ecinteractive script from hpc to your local machine:

    Code Block
    languagebash
    scp username@hpc-login:/usr/local/bin/ecinteractive .

     

  • Run ecinteractive script on your local machine:

    Code Block
    languagebash
    ./ecinteractive -u username -j

    "-u" should be used to provide ECMWF username if username on your local machine doesn't match username on the Atos HPC.

  • The ecinteractive job will start, new tab will automatically open on your web browser and attach to the Jupyter server running on the HPC. If not, you will get a link which can be used on your local machine only to paste it manually into the web browser. It should look like this:

    http://localhost:.....

  • To kill your ecinteractive job from the local machine use:

    Code Block
    languagebash
    ./ecinteractive -p hpc -k -u username

     

HTTPS access

If you wish to run Juptyer Lab on HTTPS instead of plain HTTP, you may use the -J option in ecinteractive. In that case, a personal SSL certificate would be created under ~/.ssl the first time, and would be used to encrypt the HTTP traffic between your browser and the compute node.

In order to avoid browser security warnings, you may fetch the ~/.ssl/selfCA.crt certificate from the HPCF and import it into your browser as a trusted Certificate Authority. This is only needed once.

Customising your jupyter version and environment

By default, ecinteractive will start the jupyterlab coming from the default version of python 3. If you wish to customise the version of python or jupyterlab, or simply want to tailor its environment in your ecinteractive session, create the following file in your Atos HPCF HOME:

No Format
~/.ecinteractive/jupyter_setup.sh

Then add in it the commands needed to set up the environment so that the jupyter and node commands can be found in the path. This would be equivalent to the default behaviour:

No Format
module load python3 node

Examples of contents for ~/.ecinteractive/jupyter_setup.sh

Code Block
languagebash
titleUsing the newest Python
module load python3/new node/new


Code Block
languagebash
titleUsing jupyter from a custom conda environment
module load conda
conda activate myjupyterenv


...