Skip to content

Contributor-Analyst Dataflow organization

If you are using Contributor-Analyst links I recommend using a such scheme:
Create 3 libraries for each Contributor application:

    1) skl — for skeleton library, containing only Contributor (Update-packed) links
    2) exp — for exporting data from Contributor to Analyst, containing all cubes for export. Elist is substituted to a “real” one.
    3) imp — for importing data to Contributor from Analyst, containing all cubes for import. Elist is substituted to a “real” one.

Why so?
- You use this libraries as a “stub” for contributor applications.
- You can clarify export-import-internal logic by separating the libraries
- You can easily copy applications by using Change source on D-links (exp libraries posses only links from contributor, imp — to contrib application)
- By “stubbing” Contrib-Analyst interaction you can change interaction type (Links or Publish) without touching rest of model
- this trick is proven by practice :)