February 19, 2009
Microstrategy tidbits
I chatted with Microstrategy Wednesday in a call focused on the upcoming Microstrategy 9. There wasn’t a lot of technical content, but I did glean:
- In Microstrategy 9, virtual ROLAP cubes will be able to draw on multiple relational databases, not just one. (Frankly, I’ve never understood why BI vendors are so slow to put in features like that.)
- Actually, in Microstrategy 9 cubes won’t just be virtual. You’ll be able to instantiate parts of them in memory.
- The in-memory part requires manual intervention. However, that intervention can be as minor as pushing a button to accept the recommendations of a Cube Advisor.
- The Microstrategy Cube Advisor will examine workloads for a month or so to see which queries chew up the most resources.
- Another new feature is “complete” OLAP drilldown from any point in any chart or graph, without pre-programming or pre-specification.
- Microstrategy’s favorite DBMS partners are, in some order, Netezza and Teradata.
- Microstrategy 9 is currently scheduled for March 23 release.
Categories: Business intelligence, Memory-centric data management, MicroStrategy
Subscribe to our complete feed!
Comments
3 Responses to “Microstrategy tidbits”
Leave a Reply
Kurt,
I think I can answer that question about the multiple databases. 10-15 years ago when we first started with BI tools (Microstrategy, Brio, Information Advantage, etc.) it was pretty clear that if you wanted performance, you had to design the database (physical) to be compatible with the idiosyncrasies of the BI tool. That arguing against going against multiple databases and kept you pretty much tied to the data warehouse or data mart.
-NR
Neil,
IIRC, Business Objects handled multiple databases a lot sooner than other BI tools, even before the server-side development focus of the late 1990s (“Darwin”) and so on. Is my memory off?
CAM
Curt,
BO “handled” multiple databases, but it was a contrived situation. A lot of work had to be done by IT to create a report environment that looked like a multisource ad hoc environment. The way they got around the performance issue was by pre-building SQL and loading data into a multidimensional structure. Microstrategy has introduced multisource query as an extension of what they already do – no pre-building, no intermediate structure required (they do have an in-memory ROLAP now but it is not a required for multisource).
-NR
-NR