What is the best approach for tracking versioned data in workflows?

This is a general question related to the Workflow Foundation (.NET 3.5) and the version of the data it works with. We have many custom actions that work with some data, and this data can be interesting for the future analysis of already completed work processes (provided that we configure the tracking so that it stores it in serialized form).

It may be necessary to display data from the past in the user interface, but the data will inevitably change the structure (class definition / internal structure, if it is dynamic), and the redistributed version of our library will contain a new data definition, while serialized data in the tracking database will still be in the old structure.

Is it better to use dynamic structures that do not change from the very beginning (for example, a property bag), or rather, later work with redistribution and somehow convert a serialized BLOB to a new one? Have you ever used some kind of approach in a similar scenario?

+3
source share
1

, . GAC , . , , . . , , , , . .

, . . .

+1

Source: https://habr.com/ru/post/1730223/


All Articles