Easily build, customize, and reuse compute environments (Jupyter Notebooks) - Launching 1/16/26

Beth Sheets
  • Updated

What we're solving

Terra notebook users who want to customize their own compute environments have given feedback that 1) the offered base images are too large, making them time-consuming to install and manage, 2) sudo permissions are needed to install certain tools, and 3) reusability of compute environments is impacted when Terra makes updates to the offered base images. 

Users will now have the option to leverage a slim base image with only the minimum necessary packages installed, which they can customize with elevated permissions.  

Benefits to the user experience

Users can launch the slim base image quickly, and from there, install tools however they prefer, including running their reusable image inside the environment. 

Our release plan for this change

We released a new lean, base image to our available options that will grant you sudo permissions and allow you to build a custom environment quickly. The image is available at: us.gcr.io/broad-dsp-gcr-public/terra-base:1.0.0. Learn more in our documentation.

 

Next, we will update the Python image to use our new base image, which will unlock sudo permissions for cloud environments using this image as well. The new base image also provides optimized performance and will protect your environment from accidental corruption. This may change the behavior of some packages, but you will be able to install what you need with sudo permissions. 

 

Later, we will periodically release these updates for the remaining images. We encourage you to share feedback in the comments section below or with support@terra.bio as we iteratively roll out these new enhancements.

Try it out

The new base image has the minimum needed to run Jupyter and is available at: http://us.gcr.io/broad-dsp-gcr-public/terra-base:1.0.0.  When you create a new cloud environment, select "Application configuration", then "Custom Image" and paste the image URL above. You can now build off this image with a start-up script, using sudo, or as the base of your own custom image. 

 

Learn more in our documentation: Sudo: what it is and how to use it

Was this article helpful?

2 out of 2 found this helpful

Comments

0 comments

Please sign in to leave a comment.