Business Intelligence Network business intelligence resources

Blog: William McKnight

Main

January 25, 2007

Enterprise Information Integration Architecture

This blog entry is co-authored by Mike Ferguson (link) and William McKnight and is being cross-posted on our blogs.

When you’ve decided the appropriate forays into EII (staying away from the times NOT to use EII) and you’ve selected your product, you will need to architect it into your information management environment.

The process of getting the EII tool to learn about a data source is called mapping. From this exposure to the underlying sources, you can use the EII tool to create a virtual schema, which will be used on data access. All EII applications will then see and use the single virtual schema.

The technical base of the data sources can be relational databases, packaged applications, file servers, web services and potentially numerous other data stores – operational and decision support e.g. data warehouses. Indeed, this will be a major criterion of your tool selection. EII platforms differ somewhat in the data source types supported.

Generally, one of these EII instances per environment with multiple data sources is all that is necessary. If your users already use a portal to access information systems, the EII platform can become another of the underlying data stores accessed from the portal.

Applications built on the EII platform can support SQL, XQuery for XML, ODBC, JDBC, and other APIs to access the heterogeneous data sources hidden by virtual view(s) defined in the EII platform.

Continue reading "Enterprise Information Integration Architecture" »

Posted by William McKnight at 7:26 AM | | Comments (0)

January 12, 2007

Enterprise Information Integration Products to Watch

This blog entry is co-authored by Mike Ferguson (link) and William McKnight and is being cross-posted on our blogs.

If you’re ready to take the plunge into EII, you may be surprised to find that one of your existing information management vendors is in that marketplace. It’s also a marketplace of opportunity, where new, focused vendors have emerged.

There are two main types of EII vendors in the marketplace:
1. Model-driven federated query EII vendors
2. ETL tool vendors providing EII via data integration services built using traditional graphical data flows and published as web services

Federated query EII products include
o Business Objects Data Federator
o BEA AcquaLogic Data Services
o Composite Software Information Server
o Denodo
o IBM WebSphere Federation Server (formerly WebSphere Information Integrator) and IBM Information Server
o Ipedo XIP
o Metamatrix
o Sybase Avaki
o XAware

Continue reading "Enterprise Information Integration Products to Watch" »

Posted by William McKnight at 8:17 AM | | Comments (0)

January 10, 2007

When Not to Use Enterprise Information Integration

This blog entry is co-authored by Mike Ferguson (link) and William McKnight and is being cross-posted on our blogs.

Much discussion abounds about when not to use enterprise information integration (EII). This blog looks at some situations that are not particularly well-suited to the use of EII technology as a solution. Please note that these criteria should be taken only as guidelines.

Generally speaking EII is NOT suited for
o Complex transformations, fuzzy matching and integrating high volumes of data
o A replacement for data warehousing
o Business process management
o Frequent federated query processing with single federated queries integrating data across a large number of data sources. Performance may become an issue as more and more data sources are added to a data integration server. Several vendors do support caching in order to counteract this problem but nevertheless if you plan to integrate data from a wide range of data sources in a single query you would be well advised to benchmark products and compare results before making any purchase

Continue reading "When Not to Use Enterprise Information Integration" »

Posted by William McKnight at 8:54 AM | | Comments (0)

January 5, 2007

When to use Enterprise Information Integration (EII)

This blog entry is co-authored by William McKnight and Mike Ferguson (link) and is being cross-posted on our blogs. This is the first in a series of entries on Enterprise Information Integration (EII). EII is gaining traction for enabling data integration without the need for the physical instantiation of the integration. In other words, EII adds integrated reporting capabilities while minimizing impact on existing systems. We have been selectively adding EII to our data warehouse architectures. Today, we’ll look at those situations when EII makes sense for data integration requirements.

1. Connecting structured (as in data in a data warehouse) data in particular with unstructured data takes advantage of EII’s strength of leaving data in place that could dramatically increase overall storage requirements if duplicated

2. When immediate data change in response to the data view is desired (changing a copy of the data would not suffice)

3. When data transformation is relatively light or nonexistent and just getting the data together for integrated query is the biggest challenge

4. When the relatively worse query performance of EII query is acceptable (versus the obvious advantages of physically cohabitating all data for the query)

5. Some operational and regulatory reporting where the data needed is not completely integrated in one place

6. Integration of Performance Management software with multiple underlying line of business BI systems to allow a company to see performance management at the enterprise level (across line of business BI systems) using LOB metrics to calculate enterprise KPIs

Continue reading "When to use Enterprise Information Integration (EII)" »

Posted by William McKnight at 7:33 AM | | Comments (2)

May 23, 2006

EII: Is it for real?

Enterprise Infomation Integration, or EII, is now the province of a multitude of ETL vendors. EII gives the user immediate access to information in disparate data stores. However, other than some light operational message passing, it hasn't had much commercial acceptance and little impact to the data warehouse has been felt. So, at this stage, some are asking if it's going anywhere.

True, it's still immature technology. However, there are a myriad of reasons for the lack of acceptance that would be there even if the technology were mature. It's a paradigm shift in how we've managed data/information for the last 25 years. We're used to moving data where it will be used. Some believe it's the foundation of the enormous DW/BI industry (as opposed to providing business advantage with data/information in whatever architecture fits). Plus, corporations have undertaken an unprecedented number of major projects involving data recently and needed to take on proven approaches in the process.

I think the vendors are onto something here. Ascential (IBM) is focusing efforts there. Informatica has a relationship with EII provider Composite Software, a likely acquisition target by someone at some future point. Other stand-alones are Attunity, MetaMatrix, who added the former head of global markets for Gartner as CEO last month, and Certive, who added the former vice president of engineering at Hyperion Solutions to head engineering last month.

Watch this space.

Posted by William McKnight at 8:09 AM | | Comments (1)

August 30, 2005

EAI Market Growing

According to Research and Markets, the EAI market is expected to grow from $2.4B in 2004 to $8.2B in 2011.

Sun has recently completed its purchase of EAI vendor SeeBeyond. (Link)

EAI, or virtual data integration with service oriented architectures, is a natural evolution of business intelligence. It's BI without a separate instantiation of the data, as in a data warehouse.

Continue reading "EAI Market Growing" »

Posted by William McKnight at 7:06 PM | | Comments (1)