How do you evaluate ROI and Breakeven point before migrating to Cloud - IAAS model?

IaaS covers most things that exist in a data center these days, but are delivered as a service from a cloud computing provider. This includes mostly storage and compute services, but IaaS providers are expanding into other areas such as databases services. Microsoft Azure, Amazon’s AWS (Amazon Web Services), Rackspace, and GoGrid are the major IaaS players out there, but there are many upstarts as well.

There are no hard-and-fast rules that determine if an IaaS cloud will provide savings to your business. You need to consider where you are, and where you need to go. For instance, IaaS providers can typically provide storage and compute services using usage-based pricing that’s less than what you would pay for your own hardware, software, and data center space. However, if you’ve already plunked down10 million dollars on equipment and software, you may find that your sunk costs make the cloud the least economical solution.

Below are six areas to consider, in order to calculate an accurate ROI:

Storage and Compute Costs

Even if you’ve yet to purchase your own hardware and software, you need to consider the true costs of keeping infrastructure services local versus the costs of leveraging an IaaS provider. Take storage for example. One IaaS provider allows you to store up to10 GB per month at no cost, and then charges $0.15 per GB, per month, after that. Another IaaS provider charges $0.15 per GB for the first50 TB, and then $0.14 after that. Local storage may cost $0.50 per GB to purchase, but $0.60 per GB each year to maintain, just as an example. But not so fast. There are other costs and concepts to consider as well, which I’ll cover next. (The approach to compute costs is much the same; you’re just comparing cycles instead of GBs.)

As I work through these issues with my clients, it surprises me how close IaaS-delivered storage and compute services costs are when compared to costs to purchase and maintain your own hardware and software. Broken apart into GB and cycles examined over a5-year cycle, the difference is typically plus or minus15-20 percent. However, prices for storage and compute from IaaS providers are dropping, and are likely to drop significantly over the next few years—but so are hardware and software costs.

Bandwidth Requirements

Other things to consider are an IaaS provider’s bandwidth requirements. Keep in mind that data has to be constantly transferred back and forth, and that’s going to impact both your bandwidth bill as well as performance.

SLA

You also need to consider the service level agreements (SLAs), which vary greatly from IaaS provider to IaaS provider. Most provide a five-9s (99.999%) uptime promise, with lower cost services providing four nines (99.99%), and higher cost services offering six nines (99.9999%). Some “premium services” even offer ten nines (99.%). The more nines, the higher the promise of uptime, and thus the more you pay.

Cloud Benefits

Additionally, make sure to factor in the value of expandability, scalability, and agility that cloud providers offer—consider the features and functions offered by each cloud player. All claim to provide expandability and scalability, but I always prefer to test for those capabilities before assuming the value is there. Much of the ROI of IaaS is in these areas of expandability, scalability, and agility, though determining where it comes from might not be easy to figure out.

Security and Governance

Don’t forget to consider security and governance, as security for cloud-based IaaS systems is typically more costly and complex. One of the advantages of IaaS is that it’s a highly distributed approach to computing, so systems scale through distribution of processing. However, this means you’ll need to move to more sophisticated security models, such as those that leverage identity management approaches.

People

Finally, be sure to consider the cost of people and their required skill sets. If you maintain local storage and compute services, no matter if you go private cloud, leverage simple virtualization, or leverage a more traditional model, you’re going to need people around to maintain the hardware and software, and a great deal of costs come with them.

The trick to figuring out a reliable ROI estimate is to model two approaches (local and cloud), and consider all of the obvious and non-obvious costs as I’ve defined above. You always begin with the system, application, and user requirements, and work through cloud and non-cloud solutions, and then consider all costs, including existing sunk costs in money already invested.

The cloud is almost always the right choice for those who have little or no existing storage or compute services and need to expand quickly. Small businesses or startups typically fall into this category. Larger enterprises have to deal with an existing infrastructure and requirements that are more complex. Thus, the larger you are the more complex the analysis is, and cloud computing becomes a less obvious solution.

Clouds, virtualization, and data storage networks is used to enable cost reduction and stretching of resources by supporting consolidation initiatives enabling agility, flexibility, and enhanced services that can improve both top- and bottom-line business metrics.PACE(Performance,Availability,Capacity and Energy) determine the Tot.cost of Ownership and ROIInformation services/Increase efficiency and effictivness.Incurred cost as$/capacity, $ per IOP ,capacity per watt,IOP per watt.

QOS and Service levels determine response time and availability

In order to calculate Cloud ROI, then the cost of existing IT services (on-premises) vs. Cloud services needs to be compared.  To get the TOTAL cost of on-premises IT services you have to add electricity costs for keeping up your DC's plus the hardware maintenance costs (example: every5 years the servers need a refresh).