Skip to content

index.md

Latest commit

 

History

History
executable file
·
79 lines (54 loc) · 4.03 KB

index.md

File metadata and controls

executable file
·
79 lines (54 loc) · 4.03 KB

Add a session launcher to your project

A session launcher defines the environment and compute resources for a Renku session. The basic steps for creating a session launcher are:

  1. In the Sessions section of the project page, click on ➕ to add a new session launcher.

  2. Choose and configure an environment (see Creating an environment for your session below).

  3. Select the Resource class that best fits your expected computational needs.

    :::tip

    Do you need more resources than are available in RenkuLab’s public resource classes? Contact us! We can configure a custom resource pool for your needs upon demand.

    :::

Creating an environment for your session

Renku supports a variety of ways of working with sessions. To determine what kind of Renku session environment is right for you, answer the following questions:

1. What language are you working in?

→ I’m working in Python

We have multiple ways of working in Python in Renku sessions! Please continue to the next question 2. Would you like packages to be pre-installed and ready to go when you (or anyone else) launches the session?.

→ I’m working in R

If your project uses renv and has an renv.lock file, Renku can build a code-based environment with your packages pre-installed. See How to create an environment with custom packages installed.

If you do not need packages pre-installed, create a session launcher with a global environment and select the R global environment. If you need to install additional packages temporarily, see How to install packages on-the-fly in your session.

→ I’m working in another language

Working in some other language? That’s ok! You can run a wide variety of Docker images in Renku sessions! See How to use your own docker image for a Renku session

2. Would you like packages to be pre-installed and ready to go when you (or anyone else) launches the session?

→ Yes please!

If you’d like a set of custom packages to be installed and ready to go when you (or anyone else) launches a session in your project, you can take advantage of Renku’s code based environments.

With Renku code based environments, you can point Renku to a code repository that contains an environment definition file, such as environment.yml, requirements.txt, pyproject.toml, or renv.lock, and Renku will build a custom environment for your session for you!

→ If your code repository is public, see How to create an environment with custom packages installed.

→ If your code repository is private, see How to create an environment with custom packages from a private code repository.

→ Having packages pre-installed is not so important to me right now.

If you don’t need custom packages installed, you can get started quickly by simply selecting one or Renku’s pre-made global environments. When you create a session launcher, select global environment and choose one of the pre-made environments.

When you launch a session with a global environment, you can still install packages on the fly. Please note, however, that these packages will not persist once you shut down your session. You will have to re-install them after you shut down and re-launch your session. In addition, the packages you install won’t be available for anyone else who launches a session from the launcher.

If you’d like to permanently add the packages you need to your session environment, see How to create an environment with custom packages installed.