We'd like to run the visualizations in marathon-jenkins-stats automatically with each build, and make them easily accessible. Potential collab with quality team as they are interested in the visualizations as well (they said that, ideally, the data is queried from postgres, and that both team city and jenkins data can be visualized similarly)
- Loosen the parameter so that this can be run for any jenkins job, not just Marathon
- Create an entry point which queries the last job ID loaded into postgres
- Download all jobs since that point from Jenkins; compile info and load into postgres, etc.
- Add incremental load option for postgres (currently the script drops, recreates schema, etc.)
- Add appropriate indices to schema so that querying job IDs and filtering by failure state is efficient (ideally we do not transfer entire dataset to R in order to visualize...)
- Modify R script to query its dataset from Postgres, rather than from the file system. It might be nice to work with either data source for local analysis.