News Stay informed about the latest enterprise technology news and product updates.

Wholesaler turns to data integration adapters for mainframe to Oracle migration

Associated Wholesalers Inc. turned to data integration adapters from Information Builders to migrate data from its mainframe to a new Linux platform.

Talk to execs at Information Builders and they're likely to tout the user-friendliness and simplicity of WebFOCUS, the company's business intelligence (BI) suite. But that's not what won it Associated Wholesalers Inc. (AWI)'s business.

"The reporting tool itself was pretty simple to use, but the driving force was the integration adapters that allowed me to get the data anywhere," said Stephen Kane, director of application development at AWI. "That was the one thing that they had that drove us to [Information Builders]."

AWI, a cooperative grocery wholesaler based in Robesonia, Pa., supplies inventory – meat, produce, and so on -- to around 2,400 independent grocery stores throughout the Mid-Atlantic. AWI helps the stores not just manage inventory but also coordinate advertisements and product specials.

As of 2005, however, AWI "sort of had data everywhere," Kane concedes. At the time, AWI had various legacy applications and Oracle databases scattered throughout a Windows, Unix and mainframe environment. Kane wanted to retire the mainframe, transfer its operations to a Linux platform running Oracle databases with Oracle and Java applications, and deploy new BI capabilities on top of the new system. But he knew he couldn't do it overnight.

"I knew it was going to be a five-year process, and I knew I was going to have to keep the legacy data as well as the Oracle data," Kane said. "I had several different platforms, and I had to be able to keep data in sync the whole time I was doing this."

More on data integration
Find out what's helping SaaS-based data integration tools gain momentum  

Learn why Bloor Research thinks Pervasive has the best data integration platform  

Learn how to streamline data integration projects with an integration competency center
Among the vendors AWI considered to help it integrate live inventory data from the mainframe to the Linux platform were Business Objects and IBM. AWI found both lacking. "We considered moving to IBM, to their WebSphere product, but they didn't have all the integration that we needed," Kane said.

AWI needed integration adapters that could integrate transactional data from various sources – including but not limited to analytic-friendly data warehouses – in real time to a BI system to monitor inventory levels and manage ad planning. New York City-based Information Builders' iWay data integration offerings soon caught Kane's eye.

According to analyst firm Gartner, which recently named Information Builders a leader in its BI Magic Quadrant report, its integration platform for "enterprise, real-time reporting from multiple data sources with integrated ETL [and] data federation … makes it better suited for data warehouse-less and operational reporting than most other BI platforms."

"iWay Software's adapters give developers and tools a standard way to reach out to different databases, files, and legacy systems," explained Jake Freivald, vice president of corporate marketing at Information Builders.

"Most competing products have different ways of reaching out to different things," he added, "and they often require programming to work. But iWay's are all built on a single framework that completely unifies the developer's view of the information."

After a successful proof of concept, AWI was convinced and started the five-year project to migrate from a mainframe environment to a new Linux platform running Oracle and Java apps, and Information Builders' WebFOCUS reporting tools.

During the migration process, while applications were being transferred from the mainframe to Linux platform via the integration adapters, AWI also used the adapters to connect some data still on the mainframe to its Oracle databases. That way, it could realize some improved reporting capabilities before the migrations were complete.

AWI was able to integrate sales data, housed in an Oracle data warehouse, with mainframe-based order management data, for example, to create summary reports with drill-down capabilities.

About four years into the process, AWI has completed around 80% of the migration and is already enjoying much improved reporting and ad hoc analysis capabilities, Kane said.

For example, AWI's ad planning system, once mainframe-based, is now supported by an Oracle data warehouse that runs on a more flexible Java/J2EE-based Linux platform. Instead of just weekly updates, as in the past with the mainframe, AWI managers now get daily 6 a.m. WebFOCUS-based reports, plus the ability to run ad hoc reports on demand, to better manage inventory against scheduled ad circulars that its stores plan to run.

The whole ad planning process takes eight weeks, according to Les Morton, manager of Web services and order management at AWI, and, before moving off the mainframe and implementing the Oracle warehouse and WebFOCUS, it was "fraught with inaccuracies." In some cases, retailers would end up getting the wrong items "that didn't match up with the front page ad that's going out to the consumer."

"We want to provide the independent stores the ability to compete with the big chains so they can [for example] have an eight-page ad that has all these items that have hot prices and will draw the customers into their stores," Morton said.

"One of the reasons we can provide them more insight is because now with an Oracle relational database environment, the data is more normalized … we can crosscheck information in a way we couldn't do before with the mainframe," he said. "And WebFOCUS gives AWI managers an easy way to consume the data."

AWI hopes to complete its five-year plan on schedule by the end of this year, having only its purchasing system and related data left to migrate off the mainframe.

Dig Deeper on Enterprise data integration (EDI) software

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.