By: Penny Collen, Financial Solutions Architect for Cloud Cruiser, Inc
The momentum for cloud deployments is building. Most entities have some level of cloud strategy in place and are trying move workloads in an orderly manner. Just having a strategy in place doesn’t eliminate issues, but it certainly is the right place to start. The motivation to get more work done faster and cheaper is putting pressure on the CIO to evaluate and implement cloud solutions quicker.
The reality is that a successful cloud services delivery demands thorough analysis of the workload being targeted for the move. Sadly, most cloud deployments are based on emotion – they feel right. Or they feel like they will relieve some external pressure. Once the work has been deployed to a public cloud or a private instance, problems often arise. To naysayers, it is the fault of cloud technology. All that cloud stuff was just a bunch of hype. To those asked for the deployment, there are disappointments, wasted resources, and loss of customer confidence. What actions should IT undertake to increase the probability of success? The number one action – pick the right workload!
The cloud strategy should have governing principles in several areas.
- The strategy should define the desired business benefits. It’s difficult to define success if the goal is not clearly stated.
- There should be a use case and a list of core requirements that address issues like compliance and business continuity. This forms the basis for more complex assessments of security and data placement.
- Required service levels must be detailed and agreed upon. Managing expectations of performance is a critical factor in instilling confidence with users. Service levels will have two aspects, one from the provider and one from the internal operations staff. The internal staff must learn to monitor, communicate, and coordinate information between the user and the provider- a new role for many.
- There must be an architectural design to document integration points. The cloud presents new opportunities and new risks; acknowledging both and defining how the workload fits into the overall data management plan is very important. .
- Saving money is a major motivation; however, a significant number of organizations can’t say with confidence what it costs to operate their cloud workloads. They may have a general idea but they probably haven’t gathered consumption data from their current resources at a granular level. Before moving to the cloud, it’s important to get a good baseline of how much it costs to operate today’s workloads so that credible cost comparisons can be made. While the cost models currently in use by enterprises may serve well in a legacy IT environment, those models will need to be adapted for a consumption-based approach.
Resource utilization should be tracked at a granular level and mapped to the specific departments, geography, and/or applications consuming those resources. If IT is not able to report on the cost of current workloads in very specific terms, there is no sound basis of comparison when evaluating bills from providers. Lacking a defensible cost baseline, any statements showing increases or savings are suspect.
Don’t stumble blindly into cloud deployments. Take the extra steps to get a solid set of data-driven facts together first. This may be the catalyst to justify an investment in a state-of-the-art resource and financial management solution in order to collect and aggregate usage. Many organizations are concerned about understanding the provider invoice, which is only one side of the equation. The most successful companies are realizing the value of using a solution that manages the hybrid environment, public, private, and traditional IT, in a single pane of glass. With the variety of cloud options available, it is not practical to approach cloud control in a piecemeal fashion or with simple spreadsheets. It is worth making the investment in a solution early in the cycle rather than waiting until those first invoices arrive and the charges can’t be understood.
The bottom line, make time to analyze the potential workloads so they fit your strategy for operating in a hybrid environment. Understand the business requirements and select the provider which best meets those requirements. Gathering facts and data before actual deployment will greatly increase the probability of a successful cloud experience.