Skip to content

Jupyter Notebook: login

First connect to the Jupyter Notebook Web page (you may need to provide your platform credentials), then click on Jupyter Notebook server.

Step 1: Book your CPU and memory

You should then have the following page:

Jupyter Notebook startup page

The meaning of the input is:

  • Version: this is the Conda version. For the moment, only one version can be used, since we do not know how to use several Conda version concurrently.
  • Node type: there is only one possible type.
  • Wall time: this is the time (in hours) you will be allowed to use Jupyter Notebook. You will be disconnected after that.
  • Cores: this is the number of cores that will be booked. You may be tempted to use many. However, please keep in mind that there is only one node, which is shared by all the users. Please only book the CPU you will be using.
  • Memory: this is the RAM that will be booked. As previously, book the resources sparingly.

Click on "Launch" when you are done.

Step 2: Wait for the resource

You should then have the following page:

Jupyter Notebook waiting page

This means that the cluster is waiting for the resources to be free.

If you wait for a long time (more than a few minutes), it means that either:

  • the node is full, there are already many users and you will have to wait for the others to finish;
  • or you do not have the rights to access to the services (please then refer the to previous page to request for it).

Step 3: Start Jupyter Notebook

You should then have the following page:

Jupyter Notebook starting page

You can now start Jupyter Notebook by pushing the blue button: "Connect to Jupyter Notebook".

Jupyter Notebook will open in a new window.