Walking in the ‘Cloud’
Going by the hype cycle, as defined by Gartner, and my experiences from the past two years, it is clearly evident that Cloud is here to stay and will be the focused area for enterprises in this year. However, Cloud has a number of flavors, and not all varieties of Cloud will be the immediate focus for organizations. Here is my take on the rate of adoption of various cloud services.
The Infrastructure-as-a-Service (IaaS) on the public cloud is something which has come of age, and is a pretty mature offering. I feel that more and more organizations, including large enterprises will see the value in this offering and start embracing it this year. Due to the offering being public in nature, organizations will gradually hop on to it, starting with smaller temporary projects and then moving on to bigger solutions. Even for the smaller projects, most of the organizations will rely on secure connectivity between the cloud provider and their environment, either through dedicated channels, or VPN, or similar solutions.
The next thing which will see a lot of adoption will be the Software-as-a-Service (SaaS) offering. The reason for this is purely because of the niche functional area that the SaaS offering works on, combined with the reduced cost of ownership. Majority of the SaaS providers offer a niche functional solution on a shared basis to its customers, relieving them of any infrastructure and management worries, as well as providing a pay as you go model for the services. I feel that the successful SaaS offerings will be the ones where the solutions are in niche areas and not for the run-of-the-mill functions. Additionally, the ones that will benefit will be those which offer high security controls and assurances around information isolation and multi-tenancy.
In terms of Platform-As-A-Service (PaaS), there will be a limited to a very low adoption among organizations. This is due to two major reasons. The first is the sheer maturity for any PaaS platform, which is almost non-existent. Recently some of the heavyweight IT majors have introduced their own offerings around PaaS. However, almost all of the offerings are piece meals and cannot qualify as a full blown PaaS platform. Apart from them, some smaller startups have taken a shot at creating an exhaustive PaaS solution. But questions on their credibility inhibit widespread adoption of these solutions.
The second reason is the same with all other public cloud offerings, but which amplifies in case of PaaS, that is, security. In case of IaaS, organizations were using the infrastructure to host their complete solutions over the internet. Hence, the only thing to worry about was the data transfer. In the SaaS case, the product was owned by the providers, and again, the only thing to be considered was the security of the data transfer. However, in the PaaS case, apart from the data, one has to worry about the security of their solution source code, or their IP as well. If an organization is having complex algorithms to perform; some of its core business functions; having these algorithms on a public cloud platform to be executed, there might be a big risk, with no clear standards and control in place with respect to Cloud.Hence, PaaS adoption will be limited to trials and pilots in most of the cases, and too in order to move the non-critical and non-IP based solutions to the Platform Cloud, for the time being.
Private Cloud is something that will see the maximum interest among customers. For one, this is purely due to the concerns around security and legal. Secondly, the global economic situation is compelling IT departments to look for solutions that leverage their existing investments as well as reduce the total cost of ownership. This will result in Private Cloud and Managed Private Cloud being one of the most adoptable offerings. Here again, what will be more attractive will be Infrastructure Private Cloud (Private IaaS), as it is much easier to assess, update/replace and measure Return on Investments than the more complex Platform Private Cloud (Private PaaS).
For the Platform Private Cloud, the maturity of the offerings will be an issue here as well. Apart from that, standardization will be a crucial factor. Majority of the organization will have multiple technologies, and building a Platform Cloud would require to introduce standardization, something that is extremely difficult to comprehend and undertake for organizations. Additionally, measuring the Return on Investment (RoI) will be more complex and difficult, and often involve intangible benefits such as higher availability, better reliability and increased utilization. Also, on the investment side, you will need to consider standardization costs, migration costs, as well as cost for re-architecting the applications to the platform cloud reference frameworks. Since a part of these benefits are easily achievable and measurable in an infrastructure cloud, Platform Cloud adoption will be slow. Forward thinking organizations may initiate quick pilots and Proof-Of-Concepts to evaluate and ensure that they are well-equipped to adopt Platform Cloud once it matures.
Things to note is that I am not suggesting that the offerings that will see limited adoption will be a failure, or will not see success. PaaS and Platform Cloud both may be more beneficial in the long run than other forms of cloud. It is just that, given the current maturity status of the offerings as well as the effort involved to get on board, will mean that there will be limited adoption for both of these offerings.
Yusuf Rangwala is a Principal Architect with the Services Transformation Group in Syntel. He leads the Azure Practice and has over 14 years of experience in providing Microsoft based IT solutions.