We need statistical analysis and key performance indicators to describe the lifetime of maintenance and release requests in IBS, for example:
How long does an update take starting from the time the first maintenance request is created and stopping at the time the update is released to customers?
How long does it take for a maintenance request to pass all required reviews?
How often do reviews succeed or fail, respectively?
How long does it take to create a release request after a maintenance incident has been created?
How do these values look, on average, across different code streams?
What is the absolute number of requests opened (or closed) in a given time period (per code stream)?
This information is particularly important for the Emergency Updater Team, which defines success and failure partly through the time it takes to release an update to customers.
Almost all of that information is available in IBS (OBS) in one way or another, but it is hard to retrieve the data in a format that's easy to analyze.
The aim of this Hackweek project is to remedy that issue by writing one or several tools that query IBS/OBS and generate a relational database in Sqlite3 or MySQL that contains all necessary information for further analysis with standard tools, like R, scripting languages, or spread sheets in a simple format (schema). The focus of this effort would be to support build.suse.de (IBS), but we assume that supporting build.opensuse.org (OBS) as well will be feasible, if not even trivial.
There is a relevant FATE issue at https://fate.suse.com/319971 that we'll need to take into consideration.
This project was initially discussed during the MaintSec 2016 workshop, and most of the original notes are still available at https://etherpad.nue.suse.com/p/TeamSaturn. Since then, some research and experimentation has taken place under the umbrella of https://redmine.nue.suse.com/issues/5785 and https://redmine.nue.suse.com/issues/4276.
Looking for mad skills in:
This project is part of:
Hack Week 15