June 8, 2009
Data Warehousing at the Web Layer
Over the years, we have seen numerous vendors fight for internal source data to provide business intelligence reporting. Vendors aggressively fight for financial, marketing, sales, click stream and customer data to deliver intelligence to the customer faster, cheaper and better than the others. Messaging to the industry of claims of vertical focus seem true, but in the end, are ways to manipulate customers. There has got to be a better approach...and yes, I came across one.
A few weeks ago I had the pleasure of speaking with Susan Davis, VP of Marketing and Victoria Eastwood, VP of Eng. On the surface, it seemed to be another DW database startup being thrown into the fray. Columnar database, scales to 30TB , compression, commodity hardware, yawn, yawn, yawn. However, they surprised me with the sharp left turn the conversation took. Web services, SOA, MySQL front-end, great concurrency. I believe the MySQL component of the solution is very relevant for smaller scale bi/dw, but providing a DW solution at the web tier could to be very powerful. Providing analytics capabilities for a significant amount of users and/or customers might seem to costly, but Infobright might be able to deliver it. I could easily see a MySQL datamart, powered by Infobright, utilizing real-time data warehousing techniques being brought to the front lines of self service portals at a reasonable price point. I do understand there are other ways to skin the cat, however, I could easily see this type of approach also tighten up the corporate data architectures, mdm and value-added services approaches. I think the opportunity of this approach to be endless.
Maybe I'm all wet, but do others see the power of this approach as well?
February 20, 2009
Am I buying a DW Appliance or a refrigerator?
After spending a few hours this week talking to a few Fortune 100 companies about DW Appliances and database technologies, the title of this post came to me. As these companies asked me which technologies were right for them, I had to chuckle. As I pondered the question, kitchen appliances kept jumping into my head. Call me crazy, but I believe the DW Appliance vendors have been copying the marketing brochures of Kitchenaid, GE, Subzero and Maytag and substituting ice/water dispenser for massively parallel processing / scalability, freezes ice cubes faster for executes queries faster, has a vegetable crisper for crisp analytics, and so on. At least they are both consistent with the "green" message and you plug it in and you can start using it immediately.
Like any good consultant, I responded, "That depends". I continued by asking for more information about their environment, their business objectives and their business' future needs and desires. The conversations were insightful and warrant further discussions that I look forward to.
From this exercise, I find it interesting that the marketing of the masses has had this great of an influence on companies taking their initial steps of looking for their next Data Warehouse technologies. Don't get me wrong, DW Appliances are great and have come a long way over the years. DW Appliances definitely make life easier in many different ways. Maybe a different way to look at it, has data warehouse appliances and technologies progressed to a point that selecting one is a personal preference. Hmmm...Iím not going to answer that.
At the end of the day, I hope everyone is selecting the technologies that best fit their situation. Push the marketing fluff and technology bling aside and focus on the business objectives and goals that must be satisfied first. When I say "business", I don't mean IT's perception of the business objectives, but the business objectives and goals stated in a measurable manner, by the business, to be utilized by IT to meet the goals and objectives. Focus on corporate success measurements, not IT goals when taking your first steps. Understanding the current and future business objectives will help when researching and evaluating technologies. Taking these initial steps will help narrow the field and save a lot of time while evaluating what is the correct technology for the situation.
Over the past 5 years I've done a lot of work gathering business requirements and utilizing them for technology evaluations and architecting solutions. I'd love to hear good and bad (lessons learned) experiences people have had when starting this process. Feel free to comment.
February 17, 2009
Yet another DW Appliance Vendor?
After listening to, reading and contributing to a significant number of DW technology and appliance product marketing presentations over the years, they all begin to sound the same. As I was listening to one last week, at times I felt like I was listening to Charlie Brown's teacher over my speaker phone. I don't want to name names, Oracle Exadata, but I'm sure there was greatness and significant business value within the presentation that I missed.
The interesting tidbit that came out of the Oracle Exadata webinar: Oracle is now sticking their toe into the DW Appliances pool. That is a significant milestone in the industry. Oracle now has a Data Warehouse Appliance or a pre-configured solution or whatever you want to call it. Oracle now has something to combat the 12 vendors that have had their guns aimed at Oracle's significant DW market share for years. But, the same marketing message as the others': MPP, shared nothing, no tuning, easily installed, fast, etc, etc, etc.
Wait a minute! The same marketing message? They can't do that! ...Hey Moe! I think they did.
A few weeks ago I had the privilege of spending 5 days at Oracle's HQ learning about Exadata. This is an interesting solution that the competition will have to take note of. What makes it so interesting is the upfront parallelism through RAC and the backend parallelism installed on their storage nodes. Each storage node is an HP hybrid server (cpu, mem, and disk) and has a subset of the data and processing power to run queries local to the disk. Some might argue that the parallelism is minimal, but it seems to a be an interesting first step for Oracle. Being an ex-Thinking Machines guy, it isn't exactly what I envisioned being Oracle's first step into parallelism, but it is a significant first step into the DW appliance market.
Things are going to get interesting in the data warehousing database/appliance technology market. I can't wait for the new marketing messaging from current DW appliance and technology vendors, the new trinkets and the new color of mud that will be thrown. I'm sure there will be quite the stir at TDWI in Vegas next week which I look forward to.