Cloud Implementations: Addressing the 'Gotchas'By Guest Author | Posted 01-07-2015
By Katharine Rudd
CIOs are recognizing that cloud-based applications can be successfully implemented in the context of existing infrastructure and systems. In other words, while it's certainly an innovative technology, cloud computing doesn’t mean an entirely new way of doing things. This realization has driven a pragmatic and practical approach to cloud initiatives.
For example, more and more enterprises are taking a selective mix-and-match approach to the cloud, using the public cloud for non-core software-as-a-service (SaaS) applications and maintaining private cloud solutions for more strategic applications and infrastructure. Others are recognizing that cloud benefits are often more about modernization and functionality than cost reduction. CIOs are also becoming more adept at addressing security risks, and are acknowledging that effective management processes are more important than the technology per se.
Despite this increased level of savviness, however, CIOs are still vulnerable to some surprises when it comes to cloud implementations. Anticipating and addressing these “gotchas” prior to making significant investments in cloud applications or infrastructure can yield significant benefits—and prevent major headaches—down the road.
Building a business case: One imperative is to build a business case that resonates with the CFO. A challenge here is that the transition from managing a facility with fixed expenses and relatively steady operations costs to a consumption-based model with variable pricing requires different approaches to costing and contracting, as well as effective use case analyses to predict peaks and valleys in demand for resources.
A typical cloud migration (public/hybrid/private) requires a shift from a traditional fixed and CAPEX mix to a heavier OPEX budget. Moreover, cloud initiatives require significant investments in new technology. And while the agility and flexibility benefits of cloud are significant, they don’t lend themselves to a straightforward bottom-line calculation.
As cloud and variable cost solutions continue to gain momentum, CIOs need to explain to finance executives that budgetary priorities are shifting, with declining network infrastructure costs and proportional increases in other areas.
Perhaps most importantly, cloud-based delivery means a move from fixed to variable utilization-based pricing. This introduces the unpredictability of expenditures into the equation, which can make CFOs uncomfortable. Rather than go on the defensive to justify variable costs, CIOs can make the argument that utilization-based pricing can be a positive and powerful lever to modify behavior and manage demand for IT resources.
Specifically, in a fixed-cost environment, the business has no incentive to reduce consumption, since costs don’t fluctuate regardless of the amount of resources used. A usage-based model, meanwhile, enables the business to see the consequences of its behavior and to potentially modify that behavior to use less—and therefore reduce costs.
Having the right skills: Cloud implementations involve changes in infrastructure that require new and different skill types. Even when the implementation is managed by a provider, in-house skills must be enhanced. Much of the shift involves understanding how to manage an outsourced infrastructure or software. This requires knowledge of the provider and its offerings, as well as ways to manage service-level agreements (SLAs) that are tied to the performance of someone else’s platform or system.
Additionally, rather than managing its own business customers, IT must execute a delicate handoff to a new set of managed providers. Levels of visibility will change, and, in a hybrid cloud environment, the number of parties will potentially increase. This will create a challenge for both IT and procurement.
Ensuring adequate infrastructure: Without the underlying infrastructure in place, no cloud solution will perform well. Cloud-based delivery means that many new types of traffic will come through the network, in addition to traditional data, voice and video.
This increase in traffic types, volumes and vendors is a new undertaking for many enterprises, and many lack insight into the unknown risks and potential opportunities. The result can be underperformance or even overpaying for bandwidth.
For example, moving applications such as email, ERP or big data to the public cloud will create congestion points that may not have existed before. These new capacity requirements are leading many enterprises to explore capabilities such as bigger Ethernet connections or connecting directly to providers for larger SAAS-based deployments, such as Microsoft Office365.
In pursuing these options, many companies are neglecting the potential benefits of alternatives such as carrier-neutral facilities. This approach can provide a much lower cost of connection through a colocation facility that delivers secure, direct interconnection access to a wide range of cloud and hosting providers.
CIOs need to assess their network requirements from a big-picture perspective, in parallel to developing a cloud strategy or business case. Often the network “strategy” involves a band-aid approach to add cheap bandwidth (such as broadband) to solve the congestion issue. Over time, however, this creates a hodgepodge of access, security exposure and reliabilty issues.
Bottom line: While CIOs are becoming increasingly mature and pragmatic in their approach to cloud, surprises and challenges still need to be addressed.