06-16-2014 12:45 PM
What is the best practice for managing and switching between multiple versions of the same service? There are two scenarios:
- Service description is the same, but data is different: this would be achievable by adding extra data models.
- Service description is changed for new version – adding any extra data models uses the same service description. Would this require creating an additional virtual service based on the new description?
The above is a very common question for HP Service Virtualization. Keep in mind there are several ways to achieve the goal and one size doesn't fit everyone. With that said, here is how the HP SV R&D team answered the question:
1) Same SD: Your answer is correct. You should manage this by extra Data Model.
2) New version of SD (data structure has been changed)
a) It depends if both version needs to be up and running at same time:
i) If not, you can use versioning (e.g. SVN )to checkout version you want and deploy to SV server.
ii) If both needs to be running at same time, you need to create new instance.
(1) If you want to duplicate service and update SD, easiest way how to duplicate service is to use Service Template.
(2) If you have new version SD with significant changes and you are able to execute new learning, better will be to create completely new VS based on new SD.
If you are creating more service versions, I recommend to incorporate version number/name into virtual service name.