Datalab

Working in a datalab
Author

Ludovic Deneuville

1 Principle

At ENSAI, most practical sessions (labs) use a datalab.

A datalab is a cloud environment where you can:

  1. reserve resources (CPU, GPU, RAM, storage)
  2. install your software, packages, and extensions
  3. run your programs and computations
  4. release the resources after saving your code and results

It is as if you were provided with a fresh machine, completed your work, and then returned it.

2 Connecting

ImportantPrerequisites

Have your myENSAI account activated and access to your ENSAI email inbox.

Two datalabs are available to ENSAI students:

Tip

It is strongly recommended to use the SSP Cloud first, as it is more stable.

3 How to use it?

Once connected, you will arrive at the Onyxia interface.

This interface allows you to launch catalog services (Jupyter, VSCode, RStudio, Databases, etc.) in just a few clicks.

NoteLaunching a service

For example, start by launching a Jupyter-python service:

Wait a few seconds for the service to be deployed, then click:

A new tab will open.

Your working environment is now ready.

4 Usage

Caution

On a datalab, services have a limited lifetime.

Moreover, there is no service guarantee, i.e. your services may disappear at any time.

In practice, services do not disappear except in the event of an incident or after more than one week of inactivity.

It is therefore essential to save:

  • your code (see the next section: Git)
  • your data
    • by sending it to S3 storage (not covered here, see the documentation below)
    • or by downloading it manually