Caught in the Crossfire
How to Increase the Reliability of Your IT Infrastructure Using Predictive Analytics REGISTER >
As line of business teams increasingly fund everything from CRM systems to social media apps, CIOS are caught between competing business units.
In fact, as advanced analytics continues to evolve alongside the rise in-memory computing, the data management pendulum may finally be swinging back in favor of the CIO.
Many CIOs have been trying to unify business processes via application integration projects that make sure there is a single truth manifesting itself across every business application. Conflicting data from different applications, however, can easily wind up rendering those applications all but useless. In addition, organizations often wind up having to deploy analytic applications against data warehouses that add a lot more complexity and expense.
Simo Said, vice president of database and technology for lines of business and industries for SAP, says unifying all the raw data in one system is one of the major business benefits of investing in a system such as the SAP High Performance Analytic Appliance (HANA). In effect, HANA becomes the system of record for processing all data in real time, which can then be accessed by any application that either runs directly on top of HANA or via an application programming interface. Rather than having multiple copies of the same data running throughout the enterprise, HANA provides an opportunity to centralize all that data in-memory in a way that makes distributing multiple copies of data around the enterprise unnecessary.
SAP HANA is not the only database platform that IT organizations can opt to process data in memory in real time. Exasol, for example, makes it possible to run existing SQL applications on top of an in-memory database. Conversely, Starcounter has developed a NoSQL database that runs completely in-memory.
Alternatively, some organizations may simply get by with Apache Hadoop implementation, which may not be as fast as an in-memory database, but at the very least allows IT organizations to more cost effectively build big data analytics applications.
Obviously, it will take some time for the transition to in-memory computing to fully play out across the enterprise. In the meantime, CIOs, in the absence of a corporate mandate that identifies one particular application as the system of record within the organization, may want to bide their time.
The simple fact is that a lot of the division that exists in many organizations winds up being reflected in the IT systems that support business units that are either not quite as closely aligned as they might be, or are hostile to each other because the underlying business model is fundamentally flawed.
Of course, that doesn’t stop anybody on the business side from accusing IT of not understanding the needs of their business unit, even though those needs are often in direct conflict with the needs of another part of the business. More often than not, CIOs are simply making the best of a bad business situation that is currently, but perhaps not permanently, beyond the means of their control.