Testing the Cloud

Those who have been in the IT game for some time have become accustomed to “hype” topics that all too quickly evaporate once we discover that they cannot solve ALL of our problems. While the catchy name of “cloud computing” smacks of yet another hype topic, for once, the hype is matching the reality.

One of the reasons for this is that cloud computing is nothing new, but builds on the well-established hosted services business of the last many years. What’s different about cloud computing is the scope and ambition for what can be hosted, causing enterprises to reconsider entrenched perceptions of IT best practice.

With hosted services, the emphasis was on providing remote access to applications running at a data center consolidating many applications for many customers. This provided access to economies of scale that would be otherwise difficult to achieve for individual customers.

This “software-as-a-service” model has been extended to include “platforms-as-a-service” and “infrastructure-as-a-service” to round-out the concept of cloud computing. This concept promises to provide organizations with access to all the IT resources they need to support their business without the cost and complexity of building and maintaining the IT infrastructure to maintain them.

Nevertheless, for Cloud Computing to fully deliver on its promise, it is important that the vital transition is made from the operation of hosting data to the business of providing cloud services. This is partly due to the success of cloud computing to date. As cloud computing service providers overtake more of the responsibility for customers’ data, applications and infrastructure, they also become more critical to the successful operation of the customers’ business.

In this regard, service availability is absolutely key. With cloud computing we are proposing to customers that they should abandon efforts to keep up with IT developments and outsource the operation and maintenance of IT infrastructure and services to cloud computing service providers. Without assurance of service availability and a trust on the part of the customer that these assurances will be met, this leap of faith will be difficult to take. An even worse scenario is if the leap of faith is taken and the necessary measures to assure service availability have not been implemented, leading to a very disappointed customer that is probably lost for quite a long time. Not to mention everyone that refers to this customer for their cloud computing experiences and anecdotes!

Testing and assuring cloud services, or service assurance, is therefore essential for ensuring that cloud computing fully delivers on its promise.

Nevertheless, this cannot be provided at the expense of operational efficiency. One of the motivations for outsourcing IT operations and maintenance is the economies of scale and efficiencies that large cloud computing data centers can provide. In other words, it should be cheaper per bit for a cloud computing data center to host the cloud service than individual customers implementing this infrastructure themselves.

Therefore, service efficiency is also essential and should be a foundation for all implemented cloud services.

To assure service delivery efficiently, cloud service providers need to think beyond the walls of data center and adopt an end-to-end view of the service. In the vast majority of cases, cloud service providers will need to rely on third parties in assuring services end-to-end. Building a test and monitoring infrastructure is, therefore, an essential tool for ensuring that the real-time data is available to react quickly to degradation in cloud service performance and to ensure that resources are being used efficiently.