The cloud has become the focal point of many technology discussions as businesses learn of advantages like new software licensing models, as well as much easier infrastructure provisioning. However, many of those cloud-centric discussions are light on the details regarding the hardware. As ZDNet contributor Joe McKendrick aptly pointed out, just because the infrastructure goes unseen by the customer doesn’t mean it isn’t there. In fact, cloud hardware requires many of the same considerations as the enterprise data center.
“Applications and services don’t just materialize out of the atmosphere; they, too, need to be supported by hardware and software that is built, maintained and provisioned,” Joe McKendrick. “And – very important – people to run it, monitor it, patch it, troubleshoot it, secure it, and keep things at high performance.”
McKendrick pointed to the widespread use of private cloud solutions as an example. These environments are still managed on-premises, but they’re governed using different paradigms than traditional infrastructure. The technology has necessitated many shifts – a push toward service-oriented architecture and using virtualized infrastructure rather than individual hardware components – but it is important to remember that the physical servers and storage devices are still there whether a company uses the private, public or hybrid model of cloud computing.
Although McKendrick’s article focused on how people refer to the technology, it does highlight the pressing issue of reliance on seemingly invisible hardware. Ultimately, it raises the question of whether cloud buyers are able to comprehensively understand the infrastructure that delivers their outsourced services. For cloud storage companies, addressing this issue may even be a key value proposition, as hesitant technical staff may be brought on board by a showing of hardware efficiency.
Cloud confusion
One of the advantages of fostering an environment of transparency and honing in on hardware-specific advantages is that it would help clear a lot of the confusion surrounding cloud services. Analyzing InformationWeek’s Global CIO Survey, contributor Charles Babcock offered some insight as to why customers might be frustrated by their cloud deployments. One of the common pain points is that it is difficult to truly compare how much the cloud costs to the expenses associated with doing the same thing in-house. The cost component of using the cloud can often be easily predicted because these services run on a subscription or per-use model. However, pinpointing the cost of delivering a specific service in the midsts of hardware procurement, cooling, power usage and storage costs is more difficult for customers.
Some companies in the industry have responded with initiatives to support transparency. The burgeoning interest in Facebook’s Open Commute Project is a sign that cloud storage providers are willing to open lines of communication regarding the design of their data centers. Initiatives such as OCP that place heavy emphasis on collaboration are likely to be facilitators of innovation, and similar openness could be of value to cloud providers in helping to specifically highlight of their offerings, whether it’s the ability to deliver resources efficiently or enable easily provisioned software services. Regardless of a provider’s operating model, it could be worth giving some marketing attention to the data center that forms the backbone of its services.
“Cloud data centers also are big users of open source software – Linux server operating system, Apache Web server, KVM, and Xen virtualization software – giving them a less expensive software infrastructure than the typical enterprise data center,” Babcock wrote. “Cloud centers pack in servers optimized – sometimes custom-built – for virtualization, then divide them up for multitenant use, reducing costs further. Cloud vendors spend the time to develop expertise in areas such as lower-cost virtualization software and energy-sipping server hardware because those efforts can scale even small savings across a massive installed base.”
