This article outlines recommended practices when working with cloud-based Jupyter Notebooks in Terra. For more information about the capabilities of Terra-based notebooks, explore the Visualization and Statistics section of our support knowledge base. 

1. Avoid losing data by saving output to a Google bucket

A note of caution when deleting runtimes!  Because output from analysis done in a notebook is stored in the runtime environment, it will disappear when the VM is deleted, unless you explicitly save it to permanent storage. If the current outputs are important and you want to err on the side of caution, you should save them to your local machine before proceeding with cluster re-creation.

If you are not worried about how the outputs will change after an updated runtime environment is implemented, you can re-run the notebook as soon as your runtime environment has been re-created. 

2. Re-create your cluster regularly

We recommend deleting and re-creating Notebook Runtimes every month or two to stay updated.

To keep track of how old your clusters are, you can navigate to the Notebook Runtimes section of your Terra profile:

To maintain the highest standards of OS security, and to add new features, we regularly update the implementation of the Runtime Environment. Frequent updates means you may run into trouble if a software update affects the type of runtime environment you have set up for their Notebooks. To mitigate potential problems, we recommend deleting and then re-creating runtime environments on a bi-monthly basis. Deleting and re-creating a runtime environment is quick and simple! Just follow these two easy steps:

1) click the trashcan icon in the Runtime Environment button in the upper right corner of your screen to delete:

2) Once the trashcan icon turns from blue to grey, you can click on the "play" button right next to it. All done!