CIOs should avoid ‘big-bang approach’ for SDx in 2016: Akhilesh Tuteja, KPMG

SDx is not a very specific thing and, hence, it is pretty hard to denote its market size and growth in India, as it would be similar to predicting the same for cloud services in India, says Akhilesh Tuteja, Partner and Head, IT Advisory Services, KPMG in India.

In an interview, Tuteja dwells on the different aspects of software defined and what it means for CIOs of India Inc.

How is SDx expected to be integrated into the functioning of the Indian Inc.?

SDx, in its entirety, has evolved from concepts such as Storage Defined Storage (SDS) and Software Defined Networking (SDN), etc., and it can be potentially beneficial for enterprises. However, like every potentially silver-bullet-look-alike solution, it could become difficult to enable it on high speed as we are likely to see its manifestation in different shapes, forms and positions.

Not everyone is expected to convert their entire infra to SDx as it is unlikely to take shape and size in the expected zone by most stakeholders. However, some companies are adopting bits and pieces of SDx, such as SDS which is relatively easy to deploy.

The growth of SDx is not anticipated to be superlative in India like other technologies because it is not simple to deploy.

The growth of SDx is not anticipated to be superlative in India like other technologies because it is not simple to deploy.

Which components could see more growth – SDS or SDN? And what factors may compel enterprises to invest in SDx?

These concepts are often driven by constant propagation. People selling these approaches could speak about successful cases, push very hard, and educate the market to make it popular. SDx is currently not a need which one has, but a need that has to be created. SDS is clearly a stronger proposition as it is relatively easy to deploy and, therefore, more people would want to consider it.

There is a very interesting contrast between SDx and cloud services. SDx is a concept that every company in the world could look at. The idea is that if you are on SDx then you are as good as Amazon yourself.  There is a market force pushing many companies to utilize cloud services, which means that the responsibility to deliver lies with its providers.  Majority of the cloud environments are SDx-enabled. This contradiction between cloud technologies, of giving the reins to someone else or doing it yourself (DIY) could play out in a number of areas. Having said that, SDS in my view is expected to be ahead of every other x of SDx.

And what about Indian enterprises becoming SDDCs?

We may not see software defined data centers (SDDCs) for a long time in India, unless enterprises own huge IT estate and have to upgrade their IT infrastructure. While there are advantages such as overall cost and faster provisioning time over a period of time, it could be an expensive proposition for the initial stages. It is likely that in 2016, SDDCs still remain more of an aspiration than a reality.

What according to you are the fear factors for CIOs in India to traverse ‘software defined’ path?

Let us assume that the whole world moves to cloud and there is no need for an IT estate. This is when companies requiring SDx may take to being cloud providers, likely comprising a group of not more than half a dozen powerful providers. If a CIO buys technologies to become SDX- enabled, the cloud provider can approach him or her to deliver that service instead, which is a rising trend, currently. 

From an Indian perspective, fundamental requirements for these technologies are the need for scale, large size operations and a fair bit of standardization. While most vendors promise a standardization layer at the top, typically, Indian enterprises’ stacks are fairly wide and quite heterogeneous.

Another roadblock could be the requisite skills to these technologies, because India is more of a plug-and-play market. We are not a very highly engineered system market, although that is changing now with several e-commerce companies. The evolution of e-commerce seems to be eroding the standardization of business applications’ infrastructure.

What pitfalls should CIOs and IT managers avoid with respect to SDx, amidst overhype by technology OEMs?

I would advise them to be careful of promised-interoperability versus proprietary systems. These technologies are supposed to drive optimization, but they also come with the aspect of a lock-in. I have seen that many a times, enterprises are forced to buy integration blocks, connectors (and other similar things) to make a complex provisioning work.

Another pitfall they should avoid is the ‘big-bang approach’. Although many vendors will advise to go full throttle, from the start, with SDx for more benefits, SDx needs to be implemented in small steps.

Although many vendors will advise to go full throttle, from the start, with SDx for more benefits, SDx needs to be implemented in small steps.

Lastly, companies need to invest in internal capability for these technologies. They may not have experienced resources, but they and their teams should not be novice to a technology like SDx.

Wherein the adoption curve lies for SDx in 2016 in terms of workloads or early adopter verticals?

E-commerce is a great vertical for SDx adoption; and from an industry perspective, the entire banking community. Majority of the core banking applications comprise big data and  mobility platforms. Many companies who are underinvested in technology could find more value proposition in SDx. It can be difficult to retro-fit, but it is relatively easier to build at the planning stage.

SDS being time tested, offers real benefits to companies, especially media companies as the storage needs are growing exponentially.