I have been working on a tool called HUMA – Hyperion Unused Member Analyzer. The idea for it came out of some side discussions at Kscope a couple of months ago. The idea is simple: Wouldn’t it be nice if there was an easy way to determine if any members are unused in a given cube?
Given a server, database, and credentials, HUMA will connect to a a cube, analyze its stored members, generate a list of all possible values, then iterate over it, analyzing the resulting data grids for the presence of data. If there are members with no data in them, they are shown to the user running the program. To increase performance, HUMA orders the grids and sequences of members within the sub grids so that they are aligned to the dense/sparse structure of the cube, so that it can pound on the same hot blocks before moving on to grids with different sparse permutations.
On a pretty gutless VM of mine with Essbase running in 1GB of RAM, a standard Sample/Basic cube can be ransacked for data in about three seconds. Also, given the way the tool works, it’s not necessary to do a full export of a cube or anything since the analysis is based on the data that is queried and immediately discarded. So far it seems to work pretty well.
The goal of the tool is to be a tool in the toolbox for Hyperion/Essbase admins that want to analyze their environment and act on possible improvements. This goes hand in hand with my research and efforts to find out what we all do when we dive into a new system as part of a health check hit list. Doing so on a BSO database can yield improvements (particularly on dense members).
In any case, version 1.0 of the tool is basically ready to go and I’d love to have a few people test it out and let me know of any issues!