For years, open systems disciples have prophesized the demise of the mainframe. Yet as mainframe devotees will attest, it’s unlikely the platform will disappear anytime soon.
“Our mainframe is completely reliable and dependable, and the support infrastructure is more mature than in the open systems space,” says the systems manager of a U.S.-based consulting firm. “We have numerous monitoring and audit processes in place. DB2 comprises the majority of our databases on the mainframe, and we’re able to concentrate our utilities and support processes for one database management system.”
But with open systems, he says, it’s like starting all over.
“When you have multiple operating systems and platforms with a variety of database systems, you have a more diverse database environment to manage,” he says. “This makes providing database service more challenging.”
According to IBM, more than 60 percent of the company’s current mainframe revenue comes from new workloads driven by IT trends such as Linux, virtualization, Java, and Service- Oriented Architecture (SOA). IBM estimates this trend—propelled by SOA— will cause transactions running on mainframes to easily double before 2010, and Big Blue intends to keep introducing new capabilities and environments on the platform to help today’s organizations handle the increasing volume of strategic applications and transactions being processed on the mainframe.
Still, what mainframe-based organizations do need is simplicity because let’s face it, working in a mainframe environment does pose some challenges. It’s an expensive platform to run and requires long processing windows. It uses a programming language that only a handful of IT professionals know how to operate and troubleshoot (perhaps the reason why some IT shops are turning to open systems). And there’s a limited supply of reporting tools available to extract information from the mainframe.
That said, most applications running on a mainframe are generally critical to the company’s overall function and success. Financials, manufacturing data and shipping status have to be accurate, reliable and available across departments, business units and geographic boundaries so sales order history, customer data, and the like are always available for proper reporting, trends analysis, and order fulfillment. If not, a company’s revenue and profitability could suffer.
That’s why many of today’s mainframe shops are developing a real-time data integration strategy. Real-time data integration ensures corporate data is always timely and in sync across the enterprise, providing users with a single version of accurate, reliable information so they can run an efficient company.
For instance, a bank may have critical data and legacy applications stored in mainframes that need to be integrated into other environments. A pharmaceutical company might need to integrate its legacy application with its sales application bi-directionally to deliver sales force automation. A manufacturing company might want to distribute inventory information in real-time to an Operational Data Store (ODS) so shop floor managers can react to current inventory levels. A retailer must provide continuous, real-time inventory visibility so customers can see what items are in stock for immediate online purchase.
And the list goes on. As one IT director put it: “With a real-time data integration strategy in place, we can rest assured that information residing on both our mainframe and Oracle order entry system is current and consistent. That way we can make better business decisions, fulfill orders faster, and improve customer service.”
A real-time data integration strategy captures changes to data in the databases of mainframe and other systems as they happen, flowing the data in real-time to other platforms or data stores. That’s important for companies looking to:
- Load data into Business Intelligence (BI) environments to support real-time decision-making
- Synchronize bricks and mortar, back- office data with e-business and Customer Relationship Management (CRM) data to provide employees and customers with a consistent, current view of account, inventory, and other business information
- Move data from legacy environments to open systems for new application development.
How does it work? Say an organization has in-house salespeople who are regularly entering sales order information into a mainframe as well as a large group of traveling salespeople who are simultaneously entering orders into an Oracle-based order entry application. When those working remotely connect to Oracle applications, the data has to be replicated to the DB2 sales application on the mainframe so everyone can access it and so the data between the two systems remains consistent.
With numerous mission-critical legacy applications feeding into the mainframe and no real-time data integration strategy to speak of, these remote workers can’t always be sure they’re working with the most recent transactional changes or that their sales orders are being incorporated into the corporate database in a timely manner. Accessing remotely also can put a strain on system performance, leaving internal users struggling for access to their own mainframe- based apps.
But with a real-time data integration solution in place that integrates data bi-directionally between the mainframe and Oracle environment, orders can be synchronized in real-time so there’s always a single version of accurate, reliable corporate data across the organization. When the solution is GUI-based, it’s simple to select source and target databases and then configure transformations with little programming effort. The result: faster implementation, reduced cost, and increased ROI. By incorporating a solution that runs directly on the mainframe, the data can be extracted as efficiently as possible. This allows users to continue to access business applications, with no impact on mission-critical applications such as shipping, inventory, and financials running on the mainframe.
Change Data Capture (CDC) technology comes in handy, too. Unlike traditional Extract, Transform and Load (ETL) tools, CDC captures changed data directly from database logs rather than using triggers or performing queries against the database. There’s little impact on the mainframe’s performance and there’s no need to worry about costly or risky changes to the database.
Real-time processing and replication also enables 24x7 operations. Because replication is conducted throughout the day, as the changes occur, nightly batch processing is eliminated. That means, for instance, that salespeople can access accounts and process orders no matter what time zone they’re working in. They’re more productive and have consistent customer information at their fingertips, significantly improving sales revenue and customer service.
With the proliferation of company mergers and acquisitions and the growing 24x7 world of business, companies operating on any platform face many of the same challenges. Among them is the call for continuous information access and availability across all systems and applications, whether legacy or new. With a real-time data integration strategy in place, mainframe shops in particular are achieving greater flexibility than they ever thought possible. And at the same time, they’re lowering their cost of operation, increasing productivity and customer service, and improving revenue.
Long live the mainframe! Z