Poster Abstract

P9.16 Rosa Diaz (Space Telescope Science Institute)

Theme: Evolution of software development and management

End-to-end validation framework

At ADASS XXVIII, we talked about our efforts to add scientific validation to the JWST Calibration pipeline. We have finished designing the infrastructure and have implemented many of the tests. It will consist of a series of unit test, regression tests, and validation tests that will be run automatically at a different cadence, allowing a fast development-testing-validation turnaround time and a higher level of confidence on the quality of our calibration products. The automated tests are constructed in python being a mix of pytest unit and regression tests as well as validation tests built in Jupyter notebooks. All of these test, including the notebooks, can be run automatically as part of a Jenkins testing service. Since changes to many of ground segment systems might impact our ability to produce high-quality science data, we are also supporting the development of test cases that validate all the systems; from the creation of observing proposals and execution, to the calibration and archival of the data.

In here, I will explain our end-to-end testing framework and what the Science Instruments are doing to validate the SOC system and its science products.