I have a question for my audience about a tool idea. Would it be useful to be able to tell what the data differences between two cubes with the same (or highly similar) dimensional structure is? For example, let’s say you had Sample/Basic on one server, and Sample/Basic on another server. Would it be useful to check for differences in the data loaded to them, if any?
I could see this as possible being helpful in checking for differences between cubes in development/qa/production, between archive cubes and ‘real’ cubes, and possibly during testing when you spin off a side cube to check some calcs.
Just a thought. Let me know! After HUMA is kicked over the wall I’ll be looking for my next side project (as time permits) and I am trying to focus on things that will increase the productivity of Hyperion developers.
I’m trying to develop something whether you can compare two metadata structures, already talked to Harry about stealing his xml to parent child code :).
It will be useful to have one for data differences :)
Could be interesting and useful. My only “issue” with that approach is that I really feel like generally speaking we should be managing metadata upstream (Planning, EPMA, DRM, SQL, etc.) rather than treating the cube as the source of metadata. On the other hand, though, the tool could be useful to me in the context of seeing if a dimensional structure had drifted away from another cube where they *should* be the same.
Both tools could be good for reconciliation post migration or between environments some time after migration. I guess that means they would need to be able to handle cubes on different servers
That would be really great.
I have been using a way to do this by creating a difference cube (third cube) that would XREF my Source (Cube 1) & Target (Cube 2).
We have put a formula in Scenario dimension that would find the difference. We have been using it for many years and the only difficult part is when you are retrieving on very large combinations, it takes time.
Having a seperate tool to perform this wuld be really great
Regards
Amarnath
I’ve done something similar as well, and it’s nice because it works well enough, but it’s also kind of nice to keep the cubes “clean” and have a tool that doesn’t require any setup.
Absolutely right to have a separate tool than to go with all those setup
I think this would be a very useful tool in my daily job, checking migrations between prod/qa/dev. I routinely get errors in migrating that make me think there are some differences but when the dev guy checks it he says its ok. I would be much happier if there was some kind of tool that could do the comparison like cube-delta. Has this been developed or is it still a thought?