Cloud: Decoding the real from surreal

Piyush Somani

The key highlights of 2015 were CIOs adopting a ‘Cloud First’ approach and making cloud as the focal point of their IT strategy. With cloud computing gaining rapid momentum, this year witnessed cloud moving from just being a ‘buzzword’ to making some serious inroads through real-life implementations.

To tap this opportunity, the top 4 cloud service providers have been showing inclination to completely transform into a cloud services provider. Simultaneously, CIOs of enterprises and SMEs who have been observing this transformation are under immense pressure to adopt cloud for more business agility and profitability.

In this environment, majority of the CIO’s are 'trying their luck' with the cloud implementation. Telcos, Banks, Oil & Gas companies, Shipping, Aviation and enterprises with their ERP or HANA setup have either failed to explore cloud, moved back to the traditional in-house setup or migrated to some other forms of managed IaaS. Most of them are experiencing challenges due to the top 4 cloud service providers offering “Software defined Load balancing on Virtualization” cloud model.

A quick overview into the inception reveals the functioning of this model. Cloud computing and grid computing started gaining popularity when an e-commerce giant made its in-house developed cloud platform available to other organizations. This e-commerce cloud was built on the concept of horizontal scaling on infrastructure level for catering the heavy traffic on their websites.  With this e-commerce cloud gaining popularity, other technology players continued to develop IaaS offering using horizontal scaling for other sectors. Startups and enterprises continued to change the architecture of their product and made it similar to a large e-commerce website. However, this Cloud platform was never built for the SMEs, Banks, Telcos or the enterprises as they needed more in-memory compute power along with Enterprise relational database management system(RDBMS).

Vertical auto-scaling was an ideal solution for in-memory compute processing. The core differentiating feature of vertical auto-scaling helps the virtual machines to borrow resources automatically from the cloud in real-time. In addition to hundreds of cores and terabytes of memory available in real-time to complete the compute intensive tasks, vertical auto-scaling also relieves the resources automatically once the compute demand shrinks. Vertical scaling transforms the highly commoditized business of IaaS into a utility model and successfully gave birth to “pay-per-consume” model in contrast to the “pay-as-per-use” model.

However even today major cloud services provider continue to follow this “Software defined load balancing on virtualization” model. There is limited awareness about this model being predominantly suitable to website or applications similar to a large e-commerce website. Additionally, the corresponding technology environment comprising applications and database has constantly evolved for past 20 years and technology companies have paid limited attention to evolve their cloud around this.

Last couple of years witnessed significant hype around cloud adoption. While CIOs have been talking about successful implementation on this model, the fact remains that there are very few success stories and often Enterprises have failed to go beyond the Proof of Concept(PoC). Legacy applications and databases have failed to work on this cloud platform. The key challenge is that the legacy systems need to respond to constantly evolving technologies and CIOs can only discard at their own peril.

So why is this transformation relevant today? In 2016, it is extremely important for the CIOs to embark on an IT roadmap to clearly develop their true cloud model. While making the transition, it is imperative to identify true benefits of Vertical Scaling + Scale out approach.

Additionally, Managed Services are also extremely important for the CIOs, as they cannot troubleshoot the network, operating system and database problems over Cloud. Additionally, paid managed services provided by the leading cloud service providers is through their partners, who have limited expertise of the Network and Architecture of the cloud service provider.

On the other hand, choosing a big brand for your cloud-hosting requirement may not guarantee success. A detailed technical evaluation to best suit the business requirement is critical before considering any cloud offering, as all clouds are not same.

The author is MD & CEO, ESDS