Loraine Lawson does a great job in pointing out that EII, at the least the buzzword, has fallen onto hard times.
"You probably think EII was just overhyped and petered out, right? Not exactly. EII was quickly overshadowed by ESBs, ETL and other integration solutions. But instead of disappearing, it became absorbed into niche applications and bigger products. Both companies and vendors viewed EII as a complementary, rather than a competing, solution, according to the post."
Truth-be-told EII is really about data abstraction, and data abstraction is really very important considering the value of viewing many different databases using a single virtual schema. The problem is that our business can only handle a certain number of buzzwords, so EII was renamed by many vendors, and absorbed into other products as a core component.
The reality is that the patterns of EII are alive and well, living in products sold by guys like XAware and Composite Software, but the PC term these days is "data integration." Moreover, the value of EII has not yet been replaced by ESBs and ETL, indeed they solve very different problems.
The value of EII for data integration is pretty clear to me. You have dysfunctional enterprise data out there, and placing an abstraction layers between heterogeneous enterprise data sources, and the applications that need to leverage them as a single, well defined, and well designed logical schema, just makes sense to me. Call it whatever you want to call it.
Hi David,
EII is not dead. The term/acronym was coined well ahead of organizations' ability to grok and trust the tech and "EII" fell out of vogue. It was also positioned as the replacement for (all) data integration solutions - something no single technology can achieve. Now the same tech is called Data Virtualization, Data Services or Data Abstraction (as you point out) and is a key part of a complete data integration strategy. EII lives on...even in name: www.tEIId.org.
-Ken
David -
"that which we call a rose by another name would smell as sweet." Shakespeare..
The bard wasn't the best technology guy, but he understood the value of language.
Earlier in this decade, EII was a helpful way for people to get their heads around this new class of data integration vs. ETL and EAI.
But the technology had not yet matured to the state of enterprise usage. This maturity is now in place with some enterprises deploying "eii" type solutions across dozens of projects and hundreds of data sources.
In addition, because it takes several verbs to describe EII capabilities (discover, abstract, access, virtualize, federate, deliver...etc.), where some are more important than others in certain cases, it is hard to find a one-size-fits-all term to cover this range. For example abstraction is key in SOA, whereas federation is important in BI cases.
The important thing is that with more data and complexity than ever, these verbs are getting more and more valuable.
-Bob Eve