OBIEE - Analysis Location, Usage, Interaction Clarification

Hi,

So you are talking about analysis and dashboards, and not reports, right? A report would be a BI Publisher report.

An analysis has a path, just like does a file in a file system as the catalog can be seen as similar to a file system.

All the questions you are asking is what a proper auditing tool is supposed to answer. You better forget the 100% certainty as you will never have it (there can be links defined to an analysis being part of the data directly and therefore you will never find it by looking at OBIEE).

If you look in the Catalog manager reports you can get part of the answers, but others can't be answered without looking into the XML code of every single object of your catalog.

Also keep in mind that due to security you aren't allowed, most of the time, to see what users stored in their own personal folder, and there they can have links pointing to an analysis.

So look at what catalog manager gives you and then define till where you want to push your analysis before to decide if you can get rid of an analysis or not.

Never forget that a good backup/restore strategy can help solving any issue.

Is your question answered or still open?

The lack of 100% certainty is just a fact: OBIEE is a web page in the end, which means that any piece of HTML / JS coming from anywhere can do things like a link.

It depends how "bad" were/are the guys who worked on the system till now. Many people like to do random custom things when struggling with the out of the box behaviour.

If you can guarantee that everything is in the catalog and no links/drill/navigation can come from the data, the answer is in the XML of the catalog objects. The XML doesn't lie: anything appearing on screen comes from there.

That's why catalog manager reports can give you some hints and cover the majority of cases, and the parsing of the XML itself will give the missing part. Just keep in mind special cases like shortcuts in the catalog which acts as "symbolic link" in a file system, exposing a content with path A as if it has a path B as well.

These are all nightmare cases I managed when working on audit of an OBIEE platform.

I covered the topic in https://gianniceresa.com/2018/03/gdpr-analytics-challenge/  and https://gianniceresa.com/2018/04/gdpr-analytics-solution/  (and the linked presentation https://speakerdeck.com/gianniceresa/gdpr-and-you-the-nightmare-of-ba ) and which is managed by my auditing tool using a graph database.

Cameron Loepker wrote:

 

retiring things that seem safe along with keeping backups of what was retired will be the solution.

This kind of impact analysis is precisely what the graph analytics solution does for you at the click of a button. Really one of the major selling points.