Speaker
Description
In the SKA software pipelines, artefacts from test execution contain logs (amongst others) from the pipeline which are very useful for debugging errors either from ongoing code changes or flaky tests. The success of finding the root cause for errors is typically a function of how good the logging is in the running device; however, more data points could also be beneficial. For this use case, there are four devices (at the moment) delivering a mid telescope element known as Dish LMC. The master device in Dish LMC needs to track and respond to events from the subservient devices to report the state of the dish. And having a "time machine" of events at any point enhances debugging greatly. In our integration environment, an archiver is deployed to store events from devices as they are exercised. Though helpful, it does not provide developers an immediate reference point to look when debugging issues on the staging environment. This talk demonstrates how to generate additional pipeline artefacts from device events to aid in debugging.
