User Tools

Site Tools


Remote Access to your Workstation

For security reasons, access to your Lorentz workstation is only possible within the Lorentz Institute intranet. Remote access can occur either securing your connection via an intermediate step called SSH tunneling (AKA port forwarding), through the Lorentz Institute VPN service or via the Lorentz Institute Remote Workspace.

Following are some examples that demonstrate the concept of SSH tunnelling. For alternative methods of connection, please see the relevant documentation.

Starting January 2021 you can enable two-factor authentication (2FA) on your account to further protect access to our SSH server.

:!: The examples below have been tested with OpenSSH v7.3+.

SSH tunneling

By means of an SSH tunnel you can transport any arbitrary data over an encrypted SSH connection. Members of the Lorentz Institute can use this technique to gain remote shell access to their workstation across our firewall which would prevent access otherwise.

How does it work?

You must have an ssh client installed on your personal device – e.g. laptop, PC – in order to establish a tunnelled connection.
The Lorentz Institute has a dedicated server (SSH server) ready to listen to any (authenticated) client connections.
Once a client-server connection is established, a given application contacts the SSH client on a chosen port on which the client is listening.
The SSH client in turns forwards all encrypted application data to the server which finally communicates with the actual application server.

For remote ssh connections to your IL workstation, the steps above can be summarised into the following. Establish an ssh client-server to our SSH server and instruct your SSH client to forward any new SSH-connection data that will be sent to an arbitrary port number to go via our SSH server. The server will then relay this information to the SSH server running on your workstation.

Example 1

Establish an SSH connection to via our SSH server

ssh -o ProxyCommand="ssh -W %h:%p"

:!: For connections that will use the DISPLAY environment variable (think of any application with a GUI), add the option -X to your SSH commands.

Example 2

As in Example 1 but this time using your client ssh configuration file usually located at $HOME/.ssh/config on GNU/Linux systems

# cat $HOME/.ssh/config
Host workstation
     ProxyCommand /usr/bin/ssh -W %h:%p
     User username

Once this configuration is in place, a simple ssh workstation will get you to your workstation.

Example 3

Establish a web browser connection to a jupyter notebook on port YYYY.

Configure your local $HOME/.ssh/config as below

Host styx
   LocalForward YYYY localhost:YYYY
Host workstation
   ProxyJump styx
   LocalForward YYYY localhost:YYYY

Browse to http://localhost:YYYY.

Example 4

Establish a web browser connection to a Jupyter Notebook session running on node marisXX when outside the IL intranet 1)

Host lorentz
     User username

Host maris
     ProxyJump lorentz
     User username

Host marisXX
     ProxyJump maris
     User username
     LocalForward YYYY localhost:YYYY

Browse to http://localhost:YYYY.

This method will only work if you have a slurm-controlled running jupyter session on marisXX. See xmaris.
You are strongly encouraged to use xmaris OOD facilities nonetheless.
institute_lorentz/institutelorentz_remoteaccess.txt · Last modified: 2021/01/27 14:24 by lenocil