Any realization of an infrastructure service or function that's storing data must support a portability for this data that's proportional to its volume, currency and importance to the organization.
Every realization of a service or function will eventually be phased out. Unless the data in this realization is volatile, obsolete, or unimportant, the data will have to be ported to a new realization of the service. To this end, the data must be exported from the current system, conserving as much as possible the (meta)data and correlation, and then imported in the new system, again conserving (meta)data and correlation. All (meta)data and correlation that's lost will either remain lost, or must be recreated in the new system using labour, customization and/or conversion, increasing cost (in money, time, and/or risk).
The more data portability a realization has, the less cost will be involved when changing the realization.
Data portability is especially relevant when outsourcing parts of the infrastructure. If the supplier of the infrastructure service fails, or if the quality of their service doesn't suffice, then the service may need to be outsourced to another party, or maybe even be insourced again. Lack of sufficient data portability will then present a significant additional cost.
|PAV.User Workspace.Tablet||User Workspace.Tablet||ENV.Not Specified|
No Variants preclude use of this Element (yet)