A high level look at how the disruption historical data is gathered and updated.
Attention!Currently disruption tests do not run on aggregated jobs, more fine tunning will be needed before they are enabled.
An overview for how disruption tests are implemented, the core logic that makes use of the historical data, and how to go about adding a new tests.
Job Primer ci-tool used to generate job names for Big Query.
This is an overview for how backends are queried for their availability status.