Event Details

Continous integration jobs can generate massive volumes of data. When a job fails, figuring out what went wrong can be a tedious process—one that involves investigating logs and discovering the root cause, often found in a fraction of the total job output.  By building a machine learning model from previous successful job runs, we can use the tools from the log-classify project to automatically extract anomalies from the logs of a particular run.

Similarly this principle can be applied to other use-cases as well like for example extracting anomalies from journald or other system wide regular log files. In this presentation additional use cases are covered and being discussed.