Cisco

Uresh Vahalia

The demands on computing environments have dramatically changed over the last several years.  Going back fifteen years ago, computing was easy – you had rack servers with an operating system and you ran an application on them.  Over the years, applications proliferated.  You had (and still have) mission-critical business applications, virtualized IT applications, web-based applications – each of which may have an affinity to a particular type of computing option.

Today, IT departments are faced with new challenges.  They are being asked to refresh and deploy business applications speed and consistency.  Applications and capabilities often require scale, so capacity must be easily and quickly added.  And finally, the time to move applications into production has shrunk.  This in turn has driven new requirements of IT infrastructure, including reduced complexity, single system operation, and accelerated implementation.

To meet these requirements, a new way of consuming IT infrastructure has emerged – integrated infrastructure.   Called by many names, including converged infrastructures, integrated systems, and others, these solutions brought together computing, storage, networking and a unified management paradigm that delivered fundamental capabilities:  they were pre-validated, so they mitigated risk and reduced deployment errors, they combined disciplines previously siloed, and they took the guesswork out not just deploying the infrastructure, but the hypervisor and applications that ran on top of them.

Integrated infrastructure solutions are typically delivered in one of two ways.  The first is a pre-integrated, pre-packaged system, shipped from the factory floor to the customers’ loading dock. This standardization of infrastructure is a key benefit for many companies deploying business applications, but also to standardize private cloud services.  The second is a reference architecture, which, as the name implies, provides a design reference point but considerably more flexibility.  The flexibility of the reference architecture model allowed companies to more incrementally scale up capabilities as needed. 

Typically, integrated infrastructure solutions focus on three key areas.  The first is around applications.   Most business applications, such as SAP or Oracle, are the best candidates for integrated infrastructure solutions as their requirements are usually well-known and well optimized to centralized storage.  These applications are going to be in place for a long time, so a stable, deterministic and validated solution is critical.  Many solutions will include validation and documented solutions for Oracle, SAP, Microsoft and desktop virtualization applications, as well as many vertical applications as well.

The second major focus area is around an ecosystem of partners.   While all vendors deliver a core set of technologies for integrated infrastructure, we also understand the importance of partnering with a wide range of storage, automation, management and delivery (such as system integrators or channel partners).  This provides customers with choice.  Many other vendors deliver complete vertical stacks which force the customer to standardize on one vendor. 

The third focus area is around operations.  There is often a key benefit to integrated infrastructure that is often underappreciated.  The very nature of the integration of computing, storage and networking means that all of these groups, who have deep vertical knowledge of their technology, must now come together as a team to make the best use of this infrastructure.  By using these capabilities, we have customers who have seen an 84% reduction in deployment times, a 75% reduction in cabling costs and a 61% reduction in management costs (meaning that staff can now be applied to newer, more strategic projects.

As a result of benefits such as these, you have seen significant adoption of integrated infrastructures.  According to IDC, the market is slated to grow from $2 billion in 2012 to over $7.5 billion by the end of 2015.  From 2014 to 2015, the year-over-year growth is expected to be nearly 47%.  More companies are consuming infrastructure in an integrated fashion instead of a do-it-yourself model.

But what’s coming next?  Are integrated infrastructure solutions good for everything?  The answer, not surprisingly, is no.  As stated earlier, integrated infrastructure solutions work extremely well for well-established applications.  But what about newer application types, such as big data or agile development? 

This will drive a new proliferation of computing options for IT to choose from.  The risk is that these options in turn drive silos.  After all, shouldn’t an integrated infrastructure running my SAP application be run on infrastructure that will deliver the needed performance and availability profile and same goes for my big data and analytics solution running on rack based servers. If the answer is yes, then costs and complexity will continue to increase. In the future, IT teams should consider how to extend the management capabilities of what they already have into new computing options.