Open
Description
Since we're avoiding full automation of the 10k multi-region scaletesting system, we need internal documentation and a GitHub Action for running it. This will be pretty pricy test that we don't expect customers to run, so no public documentation is needed.
With this guide, a core coder engineer should be able to conduct the test end-to-end.
Steps to document
- Spinning up the workspace infrastructure
- Starting load generation
- How to monitor and prevent failures during the run
- What to do in the case of failure
- How to access prometheus data
- How to convert data into meaningful value