Collection of tools for working with JupyterHub and notebooks.
In order to support high load events we have tooling to run stress tests on our JupyterHub deployment.
- hub-stress-test: This script allows scaling up hundreds of fake users and notebook servers (pods) at once against a target JupyterHub cluster to see how it responds to sudden load, like event users signing on at the beginning of the event. It also allows for scaling up and having a steady state of many users to profile the performance of the hub. See Hub Stress Testing for more details.
There are various configuration settings you can modify to improve both steady-state and scale-up performance. See Configuration settings for more details.
Performance data can be collected during normal operations or a stress-test run. See Profiling for more details.