About the service¶
By contrast with the compute clusters, to which batch jobs must be submitted, the machines listed here are ‘interactive’ machines, meaning that users are expected to use them directly rather than through a scheduler. The main advantage is that there is no need to wait for a job to start, but at the same time, if many users are busy on the same machine, the performances quickly degrade.
Those machines are typically used either through SSH, through a dedicated web interface (e.g. RStudio), or through a specific client (e.g. SAS Enterprise Guide.) When accessed through SSH, using a terminal multiplexer such as Screen or tmux is often useful for long-running processes that need to survive logging off the machine.
The machines SMCS1&2 are both dedicated to statistical applications. They are managed in collaboration with SMCS and offer the following services:
With 32 CPU cores at 3.5GHz and 192GB of memory, they can perform larger statistical analysis than any laptop on the market would be able to handle.
As they share a single home space for user files, you can use both ; the files you copy on one will be visible on the other automatically. That shared storage is backed up every day.
To have access to those machines, you will need to tick the box ‘Statistical computing’ when you create your CISM account.
Brufence and CeSAM¶
Several versions of Matlab are installed, which can be selected using modules. For instance, on Brufence:
$ module av matlab --- /usr/share/Modules/modulefiles --- matlab/R2010a matlab/R2015a matlab/R2013a matlab/R2017a
Beside the classical use of Matlab to perform numerical computations, the machines can also be used to compile Matlab code to standalone executables that can be run on the clusters without the need to have Matlab installed. More information on this can be found in the slides of the Matlab training session.
All three machines share a single common storage.
Access and file transfer¶
To connect to the interactive machines, you will need to use your CISM login and password.
Accessing the machines¶
On Linux and MacOS, those are installed by default. In Gnome, for instance, you can hit Ctrl+Alt+T to open a terminal window. In MacOS, search for “terminal” in the Launchpad.
Once the terminal is open, you can type in commands. To use the SSH client to connect to a machine, run the following command:
ssh -X <my_cism_login>@<machine_name>.cism.ucl.ac.be
Make sure to replace the parts in
<my_cism_login> with your actual login and
<machine_name> with the name of the computer you want to connect to. The
-X option allows using software with a graphical user interface (GUI). You can ignore it if all you need is a command line interface (CLI). Note that on MacOS, you will need to install XQuartz for this to work.
When you login, you will be asked for your password. You should then give the password you chose when you created your account. You can spare the need for typing the password at every login by using SSH keys. See Creating and using SSH keys.
On Windows, the easiest way is to install (or simply download) MobaXterm. Then in MobaXterm, create a new SSH session, and specify the machine name, you CISM login and password.
If connecting from outside the UCLouvain network, you might need to use an SSH gateway .
The usual way is to use the commands
rsync. You can always use any graphical user interface you like as long as it uses scp or sftp to make the transfers. For instance, you can run the following in a terminal:
scp -r <source_dir> <my_cism_login>@<machine_name>.cism.ucl.ac.be:<target_dir>
<source_dir> on your laptop to
<target_dir> on the machine, and
scp -r <my_cism_login>@<machine_name>.cism.ucl.ac.be:<source_dir> <target_dir>
to copy a directory back. See the scp manpage for more examples or type ‘’man scp’’ or ‘’man rsync’’ to get information about those commands.