Is hybrid cloud really the best of both?

Since its emergence more than a decade ago, much has changed in the world of Cloud services. Companies have embraced the infrastructure model, and vendors and Cloud service providers spent much of last year determining how build environments that would see businesses reap maximum benefits. Rather than the wholesale rip and replace of old networks and technology and a 100% shift to the Cloud that many forecasted, the reality has seen a more palatable and comfortable co-existence of legacy systems alongside Cloud platforms.

This mutually beneficial relationship looks likely to continue into the near future. We know from customer experience that the majority of CIOs are not in a position to build their entire application architecture, service models or even the interactions between components from scratch. Given these limitations, it’s likely the future for Cloud in the short term will be a hybrid of public and private services, probably mixed in with a bit of hybrid hosting and co-location for good measure. Ensuring that all these elements are able to not only co-exist but also integrate fully without impacting infrastructure performance will be key to the long term success of the hybrid model.

As enterprises move further into a cloud model, and their experience builds, what becomes clear is that certain Cloud models work better than others for different scenarios. While it is difficult to build out specific scenarios where one model is better than another, since each environment is unique, we can make some assumptions and draw some generalisations. Let’s take the public Cloud for example. Typically multi-tenanted, the public cloud has become the accepted norm for those who don’t require high availability or a high level of customisation and integration. The popularity of public Cloud services, such as Amazon Web Services, was created by the consumer. Whether they were at home or at work they wanted access to consumer type applications and it’s typically those types of applications that find themselves in the public Cloud.

Alternatively, the private Cloud now dominates where security and high performance is critical as is the availability of the apps the cloud is hosting. With dedicated networks, data centre processing and resources accessed within the company firewall or via a secure VPN.

But giving up control of the network and resources has proved simply unviable for many enterprises and this is why the hybrid model has gained traction this year. The trend is set to continue. According to Gartner, nearly half of enterprises will have deployed a hybrid Cloud environment by 2017.

While hybrid Cloud represents a comfortable bespoke environment and doesn’t present huge challenges technically to the IT department, the management of it still does. To ensure that you retain control where appropriate but are able to benefit from its flexibility, deployment of a hybrid solution needs to include a strategy on how to manage, control, configure and change it. As a general rule, the bigger the differences between a public and private Cloud solution the more difficult it is to manage.
In a hybrid cloud environment the network’s strength and stability is crucial and as a result, networks have grown in importance over the past couple of years. It was the network’s ability to deliver improved services and user experience that drove the popularity of the Cloud as an IT model in the first place. Given that a move to the cloud is simply about moving processing power to someone else’s data centre, the effectiveness of the network is of paramount importance, whether it’s dedicated or shared. As the enterprise extends its current domain and IP range into a third-party data centre, network design, and other key elements such as MPLS delivery, high bandwidth services, QoS, network latency reduction, will become critical in ensuring the Cloud is a viable infrastructure model.

So with hybrid Cloud continuing it’s rise in popularity, what’s next? We’re already seeing a change in the network to support Cloud, through the likes of Software Defined Networks (SDN) and application networks. We may also see a change in the way that the enterprise buys Cloud services. Rather than focus on individual Cloud service elements such as the network, computing power, storage, application or data management, enterprises will increasingly seek to outsource the entire application stack. As a result both availability and performance service levels agreements that cover the entire stack will be required for this type of environment to succeed.

We’ll also see the Cloud continue to deliver all elements of the IT environment – from storage, disaster recovery, desktop, software and infrastructure until we reach XaaS – where anything or everything is available in the Cloud. And while IaaS for commodity computing, storage and networks, and PaaS for standard applications will continue to see high rates of adoption, SaaS will gain traction as a way to support Big Data needs. We’ve also started to witness B(usiness)P(rocess)aaS as the point of Cloud interaction. This will truly provide companies with a pick and mix strategy to suits their business’s need.

Finally, we may begin to see a change to the IT department itself. As enterprises develop their own knowledge of the Cloud, their roles will move away from service deliverables and into supporting business efficiencies. Today’s network manager may well be tomorrow’s business process engineer.