Since Thomson’s question is also about the future of BODI, there is something that DNewton had pointed out in another thread. BODI and BO Data quality have been integrated (not sure if that is the right word) to form BO Data Services.
Coming to Thomson’s concerns about BODI being phased out, I feel it is a valid concern, but any ETL consultant should be tool portable and tool independent. One needs to evolve with the market changes. Senior members, please let me know if you differ.
Coming to his second question about resources on the web, I feel there are some good resources that have been already pointed out here by DNewton.
For things like vendor roadmap, opportunity planning, etc… there are a lot of blogs by some well known BI specialists which can help. For eg, this blog by Frank Dravis on Data quality, Data integration , etc is interesting. http://eimblog.businessobjects.com/dravis
The companies like Gartner and Forrester have plenty of content in which they try to predict the future of data integration technologies (ETL/ELT, EIM, EAI, etc. etc.) But you need access to that content.
But read enough of the vendor sites and blogs, and it’s a fair guess to say data integration technologies are bound to converge together. Obvisouly if one integration tool did it all – replication, ETL, messaging/queueing, etc. – and did so at a reasonable price, that vendor would probably “own” the market. But as far as I know, no one tool does it all as well as any of the specialty tools.
The other thing to keep your eye on is SOA (Service Oriented Architecture) and ESB (Enterprise Service Bus). It’s old wine in a new bottle, but the concepts are sound and those methodologies for integrating applications will eventually be the common, typical approach.
IMO this puts BO/SAP in an interesting place vs. the other big vendors (Microsoft, IBM, Oracle). All 3 of those other vendors have a strong “infrastructure” bent. You might actually go out and buy Oracle’s or IBM’s data integration or app integration tools, even if you weren’t buying the business applications that run on top of them. I don’t think this is the case with SAP. SAP is first and foremost an applications company, not an infrastructure company. So the market for the DI tool would seem to be constrained to customers that already have either the SAP or BO applications/platforms in house. Granted, that’s a big market, so there’s plenty of room for DI to thrive.
Hi Doug,
Thanks for the reply. You have raised some interesting points.
SOA, ESB in particular.
I see a lot being written about SaaS, SOA, ESB, on-demand BI, BI as a web service, etc…
Is there any real good blog / site that actually talks about the impact of these on BI / Data warehousing? Everywhere I see, I see only salesforce.com as the example. (Well, I think Nsite, crystalreports.com are something on the same line)
Do you think there is too much noise around it for now? In which case, one would do himself a favor by not reading too much about it.
Well, another area I would like to know more about it is the increasing ability to handle unstructured data like voice mail, video, etc and if there is a possibility to bring those into a warehouse?
Prem, I’m splitting this off into a new topic because I think there’s so much more we could discuss in this thread, versus the original post about the future of DI specifically.
Intelligent Enterprise and TDWI are two places you might keep looking at to cover these topics. But I agree, there’s so much noise relative to the number of companies actually using it.
I tend to look at “on-demand BI” and SaaS as implementations of service-oriented architectures. That is, SOA enables them to happen. BI web services and ESBs are other SOA technologies (implementations using SOA approaches).
I tend to ignore most SaaS stuff at the moment, simply given my own company’s tendency not to use outsourced systems, for a variety of reasons. I’m more interested in the SOA/ESB/Web services stuff.
BOBJ’s BI stack is actually well positioned to jump into the SOA world. The whole back-end of BO Enterprise uses a service-oriented approach (with some sort of custom CORBA-based framework). Moreover, BO is dipping their toes into web services: There are some new BI web services in XIr2, and it’s theoretically possible for you to call the back-end BOE web services from any other service-oriented app. Although BO hasn’t done a good job publicizing that, instead expecting you’ll use one of their “consumer” clients in Java or .NET.
What all of this will enable is more process-driven BI and Business Activity Monitoring (BAM). And better integration of BI inside operational applications. Why should analysis and reporting be separate and distinct from your transactional systems? SOA will help to bridge that gap.
ESBs are an integration tool. These tools have similar connectivity (to databases, web services, ERP and CRM systems, mainframes, message queues, etc.). But the approach is different than what DI does. At a minimum, you need things like Choreography and Orchestration to have an ESB. Check out:
From what I’ve seen customers using the most popular data integration technologies are ETL followed by EAI and then data quality. EII is trailing and product choice is difficult and a lot of architects don’t trust it for performance. The last Gartner data integration report talked about how vendors like Sybase and Informatica were struggling to sell EII licenses. Enterprise search is becoming more popular as it becomes more mature (and Google is pushing it) so it might climb this ladder.
(BOB member since 2007-06-21)
(BOB member since 2004-01-30)
(BOB member since 2007-04-14)