The resources on noto are not infinite, and limiting the cloud resources we use is important for our planet. This page regroups some practical tips to obtain the best performance while saving resources when working on noto.
See this page to get troubleshooting help in case of issue.
Avoid opening noto into multiple browser tabs
Opening noto into multiple browser tabs quickly increases the resources used. Instead, take advantage of the flexibility of the JupyterLab interface, which lets you arrange multiple documents into different tabs or panels. Panels can be arranged side by side in any way you want, by simple drag-and-drop.
You can also use keyboard shortcuts to navigate in the interface easily.
Monitor the memory you use
In addition to the baseline memory use from Jupyter, opening large data files or running memory-heavy computation can quickly make you reach the maximum RAM available to users.
You can see how much RAM is used during execution by typing the command noto_mem in a terminal on noto. To easily see which parts of your computation use memory, you can position your terminal with the noto_mem command running side by side with your notebook and observe what happens when executing each cell.
Shut down the kernels you don’t use
When you simply close a notebook, the kernel in charge of its execution remains active in the background and continues to use resources. If you open and then close multiple notebooks, the resources used quickly grow for nothing. Remember to shut down regularly the kernels you don’t use anymore.
Log out when you’re done
When you are done working with noto, if you simply close the window of your browser, the processes you have launched will continue to run in the background and consume resources until they timeout. To free up resources, remember to log out before you leave.
