As a OPCT developer I would like to make sure the release will not causing regressions on the results when comparing the results with baselines (last and older releases), so I can make sure the failures are directly related with the cluster, and not certification environment.

Goal:

• describe the steps to investigate regressions (as initially reported on SPLAT-909 )
• describe the steps to quickly compare regressions with baselines
• define a single point to store persistently the baseline results (Example S3 [currently doing by CI])
• describe the next step to automate the investigation on the release pipeline