I have a more or less simple question about some generic expectations. Here's the plot:
- several depots being spread over the country
- I've been given a "national" planning scenario with a set of com.ptvgroup.xserver.xtour.DepotSite (somewhere between 20-30)
- Furthermore there's a bunch of several thousand com.ptvgroup.xserver.xtour.PickupDeliveryOrder, each one being assigned to a unique depot
- Let's define the large national output plan, it's KPIs (number of tours, traveltime, distance) and it's calculation time as the reference level.
- From a business perspective the overall story can be split into distinct plannings per depot.
- Will the aggregated KPIs of the distinct scenarios be better than the reference KPIs?
- if we send the distinct scenarios in a single threaded sequence: will the performance be faster compared to the reference performance?
- I assume that sending the distinct scenarios being sent in parallel will return the same aggregated quality as with single threaded but the performance will be limited by the most complex one of the distinct scenarios. So this approach would return pest KPIs and performance. Would you recommend that approach in general?
- Last but not least: would you recommend to use "com.ptvgroup.xserver.xtour.OrderWithAlternativePickupDepots" for such a story?
Bernd
