September 25, 2008
Other notes on Oracle data warehousing
Obviously, the big news this week is Exadata, and its parallelization or lack thereof. But let’s not forget the rest of Oracle’s data warehousing technology.
- Frankly, I’ve come to think that disk-based OLAP cubes and materialized views are both cop-outs, indicative of a relational data warehouse architecture that can’t answer queries quickly enough straight-up. But if you disagree, then you might like Oracle’s new OLAP cube materialized views, which sound like a worthy competitor to Microsoft Analysis Services. (Further confusing things, I’ve seen reports that Oracle is increasing its commitment to Essbase, a separate MOLAP engine. I hope those are incorrect.)
- A few weeks ago, I came to realize that Oracle’s data mining database features actually mattered — perhaps not quite as much as Charlie Berger might think, but to say that is to praise with faint damns. 😉 SPSS seems to be getting large performance gains from leveraging the scoring part, and perhaps the transformation part as well. I haven’t focused on getting my details right yet, so I haven’t been writing about it. But heck, with all the other Oracle data warehousing discussion, it seems right to at least mention this part too.
Categories: Data warehousing, MOLAP, Oracle, Predictive modeling and advanced analytics
Subscribe to our complete feed!
Comments
7 Responses to “Other notes on Oracle data warehousing”
Leave a Reply
I think you’re missing at least half the point of OLAP here, and that’s the half that explains why Oracle is apparently increasing its commitment to Essbase.
It’s certainly true that one of the main drivers for adoption of ‘disk-based OLAP’ is to improve query performance, and you’re right that improving the performance of relational data warehouses will stop many people looking at OLAP in the first place. But OLAP is much more than raw performance for simple queries, it’s a rich metadata layer over your data – and this rich metadata layer in turn enables you to express the kind of complex multidimensional calculations that are simply not feasible in SQL, and to be able to do these calculations quickly. This is what MDX calculations in Analysis Services allow you to do, and what Essbase calc scripts (among other features) do too. These types of calculations are especially important in financial applications, which are of course Essbase’s traditional sweet spot, and building a high-performance multidimensional calculation engine that can handle everything a finance department can think up is by no means a trivial task. So if Oracle is indeed increasing its commitment to Essbase, it’s probably in recognition of this fact.
Chris…can you clarify re this missing metadata?
It seems to me that in a traditional ROLAP situation, there is still metadata (rich or otherwise). This information is entered into the BI tool – which keeps it in a ‘stash’ somewhere.
It’s this ‘metadata stash’ that then allows the BI tool to translate business data requests into SQL.
Chris,
It sounds as if you buy into both halves of Ted Codd’s original “OLAP” argument on behalf of Essbase:
1. Performance
2. Easier calculations
I share Richard’s question — which aspects of the metadata do you regard as particularly important?
Thanks,
CAM
Some of the critical success factors which must considered and kept in the dataware house like the mission of the business. The duration of the business are missing.
why we have to go to data ware housing
I’m hoping MDX becomes a standard, I applaud
Oracle’s 11G material cube approach but OLAP
4GL is not as elegant as MDX. MS analysis
services and reporting services can generate
full MDX queries you can tweak instead of
writing entire statements from scratch. Looking
for the next version of Essbase to do this.
I believe the power of OLAP is drag n’ drop to
minimize coding along with all the data’s permutations pre-indexed ( cube ). I wonder are
there any OLAP appliances out there that are
specifically designed for just OLAP cubes….
[…] notable point is that SPSS is more SQL-oriented than SAS. Thus, SPSS has gotten performance benefits from Oracle’s in-database data mining technology that SAS apparently […]