I've been "hearing" all day on Twitter that Microsoft would be announcing something big at OSCON2008. Perhaps this is it:

Microsoft today announced that it intends to acquire DATAllegro, provider of breakthrough data warehouse appliances. The acquisition will extend the capabilities of Microsoft’s mission-critical data platform, making it easier and more cost effective for customers of all sizes to manage and glean insight from the ever expanding amount of data generated by and for businesses, employees and consumers.
-- Press Release "Microsoft to Acquire DATAllegro"

This is very interesting given the progress that Microsoft has made with its analytic services binding MS Office and SQL Server. Further quoting from the press release:

“Microsoft SQL Server 2008 delivers enterprise-class capabilities in business intelligence and data warehousing and the addition of the DATAllegro team and their technology will take our data platform to the highest scale of data warehousing.”
-- Ted Kummert, corporate vice president of the Data and Storage Platform Division at Microsoft

The direction for DATAllegro's data warehouse appliance is also made clear in the press release:

“DATAllegro's integration with SQL Server is the opti mal next generation solution and the acquisition by Microsoft is a great conclusion for the company.”
-- Lisa Lambert, Intel Capital managing director, Software and Solutions Group.

For those who don't know, DATAllegro is a data warehousing appliance company that utilizes "EMC® storage, Dell™ servers, Cisco® InfiniBand switches, Intel® multi-core CPUs and the Ingres® open source database".

So, whither Ingres in this acquisition? As we've written before here, Ingres is one of the earliest and strongest RDBMS products, which was absorbed by CA and then spun off again with an open source play in 2005. MS SQL Server, of course, started out as a rebranding of Sybase SQL*Server, until the partnership dissolved in the mid-1990's. Since then, MS SQL Server has been geared mostly as a workgroup and data mart server. It seems that a switch from Ingres to MS SQL Server could heavily undermine DATAllegro's business. In addition, the switchover in code to T-SQL will be a nightmare for developers. Add to that the challenges of moving from Linux to MS Windows, and from C/C++ to C# and it will take quite some time in production environments to iron out all the wrinkles.

In addition, while most seem to think that this puts Microsoft in a good position to challenge Oracle for the Enterprise Data Warehouse lead, it actually puts Microsoft directly into competition with other DW appliance vendors, such as Teradata. I truly doubt that this move will position Microsoft strongly into competition with either Oracle or Teradata, but merely marks another tactical error in Microsoft's increasingly desperate acquisition strategy to move deeper into the Enterprise on one hand, while striving to move further into the online space on the other.

More can be read at:

• VentureBeat "Microsoft Acquires DATAllegro"
• Information Week: "Microsoft to Buy DATAllegro"
• Mark Madsen in Intelligent Enterprise "What the Microsoft-DatAllegro Deal Means for Customers, Vendors and BI"

Earlier today, I met John Sichi for coffee at the Half Moon Bay Coffee Co. in the Stone Pine Center. John is also a Coastsider and very involved in open source data management & analytics. We spoke of many things: our histories, folk we know in common such as Julian Hyde and Nicholas Goodman and some Oracle alums, happenings in the open source BI world, Pentaho, JasperSoft, SpagoBI, and lots of good story telling.

Mostly though, we spoke of LucidDB, LucidEra, & metadata management. I've been asked not to blog about some of things we discussed, so I'm just going to be safe and say that I am very impressed with what LucidEra is doing in BI SaaS. I'm also looking toward including the amazing capabilities of the column-store open source LucidDB in some engagement, somewhere, as soon as I can.

Just got home from the MySQL Data Warehousing and BI Birds of a Feather gathering (BoaF). I'm tired, but my mind is on overdrive. 'Tis a great feeling.

First, I want to thank Lance Walter of Pentaho for introducing Clarise and me to the group as publishers of the OSBI Lens on Squidoo, this blog and the OSS wiki.

Clarise and I had a great conversation with Dr. Jacob Nikom of the MIT Lincoln Laboratory. The conversation ranged from the great Chimay beer that Matt Casters of Pentaho and Lead Architect of KETTLE, brought with him from Belgium, to

• Data Modeling and Relational Algebra/Theory: the integrity of the model
• Bayesian Statistics, Weibull Analysis and Tensor Calculus for mathematical modeling of complex systems [I love it when n-dimensional eigenvalues start floating in front of my eyes]
• Meeting the needs of different types of users: managers, scientists, business folk
• supplementing historical data warehouses with [near] real time data using ESB and dashboards
• data visualization of complex data sets such that the analyses and limitations can be grasped at different levels by different users
• collaboration among distributed workgroups of disparate career backgrounds and cultural pre-dispositions
• use of Second Life and other virtual worlds for collaboration and data visualization
• a calculator is to a computer [think if statment] as a flat file is to a relational database [think where clause]
• early USSR vs. British knock-offs of IBM mainframes
• Complexity as the balance of robustness and fragility

At various times in this discussion, we were joined by Sherman Wood, Director of BI at JasperSoft, and one of the legendary Mondrian developers, Julian Hyde of Pentaho and Mondrian Lead Architect, and Nicholas Goodman Director of BI for Pentaho.

And if you put Nick and dashboards and virtual worlds in the same post, then you have to mention Discoverer meets Duke Nukem.

Jakob, et al, thank you so very much for a great conversation.

The schedule for the Campus Technology 2007 conference is online; a PDF of the brochure is also available for download. In addition to our session, there are several other workshops and talks related to either BI/DW or open source solutions.

• M02 - BI and Data Warehousing: Powering Better Decisions in Higher Ed on Monday, July 30 at 8:30 a.m. by Ora Fish
• Th06 - Leading Community Source Initiatives on Thursday, August 02 at 9:45 a.m., a panel led by Lee Belarmino, and featuring Rob Abel, Joel Smith and John F. (Barry) Walsh.
• T04 - Jump-Start Your BI and DW Initiatives: 60 Minutes of Solutions on Tuesday, July 31 at 10:00 a.m. by Ora Fish
• W08 - The Economics of BI: How to Drive Cost-Effective Implementation Strategies on Wednesday, August 01 at 11:15 a.m. by Joseph A. di Paolantonio and Clarise Z. Doval Santos
• W09 - What You Need to Know About Open Source Governance in Higher Education on Wednesday, August 01 at 11:15 a.m. by Chris Coppola
• Th05 - The Next Wave of Open Source: Kuali Spells S-I-S on Thursday, August 02 at 8:30 a.m. by Jeffrey C. Huskamp

We're hoping to coordinate with other speakers, so that our sessions are complementary and to avoid duplication.

Will you be going to CT 2007? What would you like to see discussed in terms of BI/DW and open source solutions? See you there.

Mary Grush has invited Clarise and I me to speak at the Campus Technology 2007 conference to be held in Washington, D.C., USA. We'll be speaking on Wednesday, 2007 August 1 at 11:15am-12:15pm. In general, institutes of higher learning are only beginning to explore data warehousing and business intelligence technologies, and, in general, they don't like what they're seeing from traditional, proprietary vendors. From our initial conversations with Mary, here's our direction. We'll develop this here in the OSS Blog as much as we can. We would really appreciate any comments to help us refine our talk.

Cost Effective BI/DW Strategy

Abstract

Our strategy for reporting, data management and analysis programs and projects responds to user needs quickly without blowing the budget. Using open source software, project management, and user involvement, this strategy economically and efficiently meets campus-wide and departmental data warehouse, data mart, and business intelligence needs through dashboards, reporting, OLAP, and data mining tools. Cost effective results can be in user's hands in as little as one week.

Points to be covered

1. A framework leading to an economical strategy/BI-roadmap for data warehousing, data management or data analytics programs
2. Program, Project & Risk Management methods
3. Risk and advantages of using open Source Solutions for BI suites or DW/data mining components such as ETL/EAI/ESB, RDBMS & MDDB, meta data management, reporting, OLAP engines, multi-variate analysis (a.k.a. "slice & dice"), machine learning, portals, and dashboards
4. User involvement for determining specifications and implementing quality control
5. Costing, value and return

Take-away Points

1. Strategy and tactics should be separated with a clear iteration plan for quick, economical response to user & organizational needs
2. Agile development doesn't mean a lack of project management nor should it allow scope creep
3. Open source software has matured to the point where it can certainly be used for prototyping and even production

Over the past five years, our research into open source BI components has shown few projects supporting BI, and no BI suites, until recently. Bee is the oldest of the open source BI suites, starting in 2002. Five years ago, there was one open source project developing an Extract, Transform and Load (ETL) tool - Jetstream, one for reporting - JasperReports, one for analysis - Mondrian. Of the open source Relational Database Management Systems (RDBMS), none were optimized for very large databases, or for querying, until this year. There are now over 25 open source projects supporting every aspect of BI, from ETL to the user Portal, including reporting, on-line analytical processing (OLAP), advanced analytics and data mining, workflow, and dashboards. Six of these can be considered BI suites, with all but Bee having launched this year.

Vladimir Stojanovski has written a five-part article in his blog at ITtoolbox. Part of his conclusion is quoted below.

Call me shortsighted, but then this nomer could also apply to the CRM/BI industry indiscriminately (except for the brave souls at places like SugarCRM [see post Open Source: CRM and Business Intelligence (Part 2 - SugarCRM)] and Pentaho [see post Open Source: CRM and Business Intelligence (Part 3 - Pentaho, et al)]). The industry is finally being forced to take Open Source seriously not necessarily because we think it is a great movement, but because our clients are forcing us to do so. An increasing number of companies are adopting Open Source in fundamental areas such as operating systems (Linux), database platforms (MySQL, PostgreSQL), application servers (JBoss), and web servers (Apache). This foundational platform is then forcing itself onto enterprise-class applications, such as CRM.
-- Open Source: Closing thoughts, I think... (Part 5) by Vladimir Stojanovski

As shown in my opening paragraph, the open source movement is responding to the interest in open source solutions for enterprise applications, particularly, BI. You can check out the links in the side column of this blog for a list of open source BI suites and tools being developed. We'll be continuing with our research and use of open source BI solutions over the past year, and I think it will be some time beyond that before we, or Vladimir, or anyone else, actually writes the final Closing Thoughts on open source BI.

Recently, Navica and InterASC teamed up on project where the customer required we use PostgreSQL as a central data warehouse. Clarise pointed out that PostgreSQL lacked essential attributes to be an efficient platform for data warehousing. In investigating alternatives, we discovered Bizgres. Bizgres is a separate distrubtion based on PostgreSQL with the primary purpose of filling exactly those lacks Clarise had highlighted such as table partitioning and bit map indexing, and the secondary purpose of building a BI suite. As we develop Open Source Business Intelligence, we'll be writing posts describing the enhancements that Bizgres is making to PostgreSQL, and compare Bizgres to Oracle as a DW platform.

Technorati Tags: Open Source, Bizgres, PostgreSQL, Business Intelligence, Oracle