Leading, Lagging And Lame BI And Data Warehousing Transformations
Lessons learned from attempts at innovation and renewal at Microsoft, Teradata, MicroStrategy and Actuate.
He not busy being born is busy dying, or so I hear, including in technology, where the rule is innovate or fail. Contrast four flavors of business intelligence/data-warehousing market transformation at segment leaders Microsoft, MicroStrategy, Teradata and Actuate.
Call the varieties "Keep up with the Joneses" (business platform leader Microsoft), "Bet the Farm" (BI-indepedent MicroStrategy), "Steady as She Goes" (enterprise data warehousing kingpin Teradata), and "Who Am I?" (BI reporting specialist and open-source leader Actuate).
Microsoft Parallel Data Warehouse
How is it that 30-plus years on, Microsoft can't shake the image among the cognoscenti of technology laggard? Never first to market, and never technically best (so far as I can tell), yet able to maintain its position by a combination of good-enough-and-sometimes-eventually-great software, strong platform integration (including with Microsoft PowerPivot), market muscle, aggressive pricing, and tenacity.
Microsoft bought DATAllegro in July 2008 with the goal of adding parallel processing to the SQL data-warehousing platform... within 18-24 months. Well, the Parallel Data Warehouse is finally here in a SQL Server 2008 Release 2 edition, announced last week at the Professional Association for SQL Server (PASS) conference, slated to ship next month. "The delay in release may reflect Microsoft's caution with a rather complex product," says Kurt Mackie, writing at RedmondMag.com.
I see here essentially no technical innovation beyond platform integration in the 28 months since Microsoft's DATAllegro acquisition. What's in the forthcoming release? As expected, Microsoft replaced DATAllegro's Ingres open-source DBMS with SQL Server and ported from Linux to Windows.
PDW is a massively parallel processing (MPP) data-warehousing appliance and will initially be available on HP hardware. DATAllegro ran on Dell; PDW will eventually come out on Dell, Bull, and IBM hardware. Frankly, I find the notion of an "appliance" with five different hardware options quite curious. The need to bring along five hardware vendors may slow development.
Columnar storage -- pioneered by Sybase IQ like 20 years ago, now available in Vertica, ParAccel, the Infobright MySQL engine, EMC Greenplum, Aster Data's nCluster, and other analytical database systems -- will have to wait for the Denali SQL Server release, annouced last week at PASS for possible late 2011 availability.
Column-oriented stores can accelerate many analytical queries, dramatically reduce disk I/O, and slash storage requirement via aggressive compression. But Microsoft was either unwilling or unable to include it in the initial PDW edition.
SQL Server 2008 Release 2 PDW also lacks in-database analytics, the ability to embed analytical computations in the DBMS, which reduces data movement, enhances security, and simplifies application development.
Most of Microsoft's analytical DBMS competitors now support in-database analytics, including ParAccel, which is set to announce in-database capabilities in its forthcoming PADB 3.0 release.
Teradata's transformation has two facets: solidifying its core enterprise data warehousing platform, including further development of a lower-end appliance line, and building out its partner ecosystem, which includes diverse BI, analytics, data-integration, and services providers. This latter step hardly makes Teradata unique among analytical DBMS vendors, but it does make Teradata notable in that, unlike most others, it has the cash and capacity to buy rather than partner.
Rather than lose focus by acquiring, for instance, companies in the text-analytics, BI visualization, reporting, or location intelligence platforms, Teradata strives to provide smaller and/or specialized partners with an ideal, high-performance analytics back-end and an entrée to enterprise accounts.
Teradata also works with large BI front-end partners that include otherwise-rivals such as Microsoft and Oracle, and also with many-partnered specialists such as Hadoop-commercializer Cloudera and BI stalwart MicroStrategy.
MicroStrategy Goes Mobile
MicroStrategy has dashboards, reporting, and dimensional analysis (a.k.a. Relational OLAP) down cold. And now the company has, from all appearances, gone mobile mad. MicroStrategy is in good shape as one of the two major BI independents (the other is Information Builders) and has bet big in a bold move to vault into a dominant mobile-BI position.
What kind of bet? (I assume payoff for company co-founder and principal owner Michael Saylor will be a bigger boat.) MicroStrategy is not only refacing its products -- check out the videos and demos -- and creating a device-suited user experience with appropriate touch and gesture interface methods. MicroStrategy is also transforming itself.
According to a September Computerworld story, MicroStrategy "has deployed 1,100 Apple iPads to executives and sales personnel to conduct critical job-related tasks. The company said it expects 700 more iPads to be deployed soon." So every one of MicroStrategy's 1,800 employees will get an iPad.
A leading BI industry analyst told me a few weeks back that MicroStrategy salespeople were told to pursue only opportunities that involve mobile, leading to grumbling that they had lost half their sales pipeline. I'm betting that most lost prospects will be back.
6 Tools to Protect Big DataMost IT teams have their conventional databases covered in terms of security and business continuity. But as we enter the era of big data, Hadoop, and NoSQL, protection schemes need to evolve. In fact, big data could drive the next big security strategy shift.
Big Data Brings Big Security ProblemsWhy should big data be more difficult to secure? In a word, variety. But the business wonít wait to use it to predict customer behavior, find correlations across disparate data sources, predict fraud or financial risk, and more.