This is a training app used to exercise higher level (non-unit) app testing. Given a vin, it will identify the vin's history, make, model and year, and then attempt to match it with existing vehicles available for sale and give a recommended price.
There are 3 collaborating services, presenting different testing challenges:
- history - cannot be injected with data, data will change over time
- mommy - make mode year. Data cannot be modified, but is stable. This service is slow.
- listings - can make and delete new listings for testing purposes. Tests from multiple workstations will collide initially
- Start up: history, listings, mommy, and vinkenstein
- navigate to http://localhost:8080/index.html
A dataset of around 35000 listings is available in the TestListings class. For most of the exercise this data will stay consistent in the history service, but at some point number of accidents and number of owners will start changing. This is a sample of real data, with vins anonymized (they will not pass check digit at this point) and prices adjusted off of their original market value, but within realistic boundaries.
