Why zBX Outperforms DIY When Deploying a Hybrid Infrastructure
Editor’s note: Dave Hayslett, Fehmina Merchant, David Rhoderick, Emily Farmer and the System z Competitive Project Office contributed to this article.
Flat budgets, demand to support evolving business objectives, and constant pressure to do more, faster have IT administrators and technology decision makers looking for any advantage. Workload optimization can help ease this burden. By matching applications with the best OS or hardware platform for specific jobs and service requirments, organizations can improve efficiency, lower operational costs and better align technology with business requirements.
One optimization approach is to reject bundled technology solutions in favor of a do-it-yourself (DIY) strategy for infrastructure evolution. After all, wouldn’t you save money by building your own customized solution from an a la carte menu of component parts? It’s possible, but unlikely. Although a DIY approach to workload optimization might deliver some measure of savings, several additional questions need to be asked when evaluating this strategy:
- How much will you really save over time?
- Have you considered aspects of a packaged solution that go beyond pricing and benchmarks?
- What are you giving up?
The Need for Smarter Computing
As our planet grows smarter and becomes more connected, new technologies frequently burst onto the scene, enabling—and often necessitating—new business models. Organizations are awash in a flood of largely unstructured data pouring in from intelligent objects and sensor grids, across social and mobile networks, through customer interactions, and from points all along the supply chain.
Forward-thinking organizations are using innovative technologies to glean insight from this data. Many are embracing the key principles of smarter computing, implementing IT infrastructures and technologies designed for data so they can leverage advanced analytics to unlock insights and drive better business outcomes. They’re creating infrastructures tuned to the task with workload-optimized systems such as the IBM zEnterprise* BladeCenter* Extension (zBX) infrastructure. Many have turned to the zEnterprise System and the zBX component to handle transactional computation and distributed computation within a single workflow. Implemented correctly, these systems enable organizations to deliver services faster, with higher quality and superior economics.
Can’t I Do It Myself?
Today’s diverse workloads often require different CPU engines and specialty appliances capable of optimized algorithmic processing and high-speed data manipulation. Some IT departments might need to run combinations of platforms for these workloads, such as z/OS*, Linux* on System z* or System x*, Microsoft* Windows*, and AIX* software.
One key issue facing DIYers is a lack of integrated network and virtualization management, which can affect efficiency and increase the time and effort staff spends on mundane chores. Combining disparate vendor technologies into a home-brewed solution can also introduce costly complications as staff scrambles to find the root cause of service or performance issues that can result from conflicts, inefficient interactions or other technological quirks. Also, poor network latency can become a serious challenge as IT struggles to make heterogeneous components communicate more efficiently.
Search our new 2013 Buyer's Guide.
Insider | In many cases, batch processing accounts for as much as 40 percent
Features | The Polaris Workshop maps out a custom platform strategy
IT Today | What to consider when adding a Linux workload to your z/VM system