In-Depth

Cloud Evolution: 5 Reasons Why 'Now' Is Good

Just starting to transition your enterprise to the cloud? Your timing might just be perfect.

• By Ian Murphy
• 07/01/2016

The first generation of cloud-based computing was all about infrastructure and cost. Companies that moved to cloud did so based on promises of lower cost from the heavy reduction in capital expenditure and the ability to flexibly increase and decrease computing power as and when required. Typical of many transitions, the gap between reality and hype caused problems, misunderstandings and probably helped delay some of the benefits companies are now beginning to see, making "now" a more optimal time to transition. Here's five reasons to consider.

1. Maturity Has Ironed Out Many Kinks
As cloud-based computing has matured, so have the models of deployment and the underlying technology. To move beyond the two extremes of all-in or all-out, issues such as performance and trust have needed to be resolved. Additional regulatory issues around where the data is stored and processed have also needed to be addressed. The solution to this has been to move away from the concept of a small number of hyperscale datacenters to smaller in-country datacenters

Perhaps the easiest issues to resolve have been those around performance. The introduction of better connectivity between business premises and cloud datacenters has improved substantially. More importantly the interconnectivity between cloud datacenters has become faster and more reliable. This means that issues such as latency in previous generations of web-based applications have receded although we are a long way from them disappearing completely.

Other issues have been addressed that has made cloud a more desirable platform for developers compared to the early days. Today's cloud environments enable fast and relatively easy adoption for developers of all levels of experience. At the same time more complex capabilities have been made available such as scriptable infrastructure for the more experience. There are marketplaces of services and greater tooling support. In some cases deployment has become one-click rather than the complicated, multi-step processes of a few years ago.

The drive to advance the security posture of cloud services and systems and raise the overall confidence of organizations has seen more secure connections between on-premises and cloud environments. Operationally the introduction of federated management tools have allowed operations teams to manage both on-premises and cloud-based deployments through a single interface. Federated security has delivered single sign-on solutions that protect users and data access enabling companies to extend processes and compliance control from on-premises to the cloud.

2. It's About the Hybrid, Stupid
One of the biggest hurdles with previous generations of technology platforms was vendor lock-in. Early cloud deployments went down the same route even those built on open standards such as OpenStack. Last year the OpenStack Foundation ordered distributors of the cloud-based operating system to unbundle proprietary components. Since then we have seen interoperability reach the point where companies are beginning to deploy clouds of clouds or, to put it another way, they have multiple cloud instances from multiple vendors and they all talk to each other and to any on-premises systems that are still in use.

Vendors offer customers the ability to create their own cloud environments that match those seen in the public cloud. This has allowed enterprises to embrace many of the benefit of a cloud infrastructure while addressing concerns such as governance, legal or regulatory compliance. These on-premises clouds can be securely connected to others that reside outside of the organization to create a hybrid cloud model. We are seeing some organizations deploy clouds of clouds whereby they mix and match different combinations of public and private and different vendor services: e.g.,  compute from one vendor and storage from another.

Those companies wanting to adopt a hybrid cloud model need the management tools to do so. Managing both on-premises and cloud environments through a single tool can be done through a number of products from vendors such as VMware (e.g., vRealize), IBM, HPE and BMC. This allows operations teams to balance where they are running their applications, move workload to where it needs to run and incorporate orchestration for faster deployment.

In effect we have reached the tipping point of hybrid cloud computing which opens up new opportunities. For some companies it is about using cloud-based applications to consume data that is held on-premises. This need to hold onto data is not driven by fear of cloud, as many were accused of in the early days, but by regulators who are concerned about privacy. A quick look at the state of privacy and lack of trust between the EU and US shows just how those regulations have come about.

3. 'Born on the Cloud' Disrupters: Paving the Way
We have begun to see the success of those early cloud visionaries: companies who decided that it wasn't enough to port their applications to the cloud. Instead they took the bold move of writing and architecting (or re-writing and re-architecting) their applications for the cloud, leading to the phrase "born on the cloud."

Companies such as Netflix or Uber took a risk. They often wrote their applications from scratch or rewrote existing code at a high cost to take advantage of the multi-tenant cloud computing model. The benefits for those companies now are that they are among the fastest growing businesses around. More to the point is that their business operations have disrupted many of the more established incumbent organizations

This sends a clear message to IT departments, many of whom are beginning to realize that cloud is not only secure but that it offers myriad business and operational benefits outside of the challenges of fully understanding and addressing data sovereignty and privacy.

The evolution of the cloud from an off-site extensible test and development platform has passed and it is now just another deployment platform. This speed with which it has reached this level of maturity has outstripped attitudes. Companies such as IBM, SAP, Oracle and Microsoft are now enabling customers to deploy production environments. Those environments are not just for applications with limited business impact but include core enterprise software applications such as ERP, CRM, big data, analytics and accounting.

4. A Changed Way of Development
The next generation of applications need to be written to run in the cloud, taking advantage of cloud architectures and capabilities.  For this to occur, it's time to put aside old methodologies and monolithic application frameworks that result in a slower pace of delivery and often more complex interdependent code bases that will certainly be more prone to errors.

To meet the demands of users for applications that can be delivered in weeks (not years), a modular, services based mentality and an agile way of execution needs to be adopted.

It is time for organizations to look at continuous development, delivery, integration and testing. DevOps is a maturing mode of collaboration and automation between development and operation that needs to be a foundation methodology for cloud delivery. 

Cloud encourages organizations to flatten out the complex development structures and adopt application programming interfaces (APIs) where software application innovation is gained through easier integration and third party development support.

Cloud helps reduce much of the complexity and improve the speed of development. For example, companies can run multiple development sites that are integrated using change management tools. These tools provide a management view over all code streams and help to simplify the integration, testing and deployment phases.

The cloud also brings a new wave of collaboration tools. IT pros not just using Web meetings to share ideas but are able to use cloud-based code repositories that resolve the problem of keeping code libraries synchronized. They are able to contribute to other developers' code in real-time and share the same tools.

5. You Don't Have To Be 'Born on the Cloud' To Live on the Cloud
The next generation of enterprise software will be written to run across a seamless estate made up of different cloud infrastructures while bridging the old and new. For example, traditional systems of record connected to Internet of Things applications through layers of APIs. Those same APIs allow IT to keep control of the data in order to meet regulatory requirements without hindering innovation. A look at the way banking and retail have transformed their once infrastructure heavy and slow to change environments shows that hybrid computing encompassing on-premises solutions and multiple cloud instances is very much here and now.

From collaborative development environments to federated management and security tools, it really is getting easier to move to the cloud. Companies are now able to use the same software in the cloud as they use on-premises. This means that they can quickly move their applications to the cloud. Moving data is more complex mainly due to the size of the data and the regulatory landscape. In a hybrid cloud environment this is not a major issue as companies can move their biggest problem, computing hungry applications, to the cloud, and access the data left on-premises.

Those companies that move first and begin to take advantage of this technology are likely to be those that survive in the long term. Compare the new generation of born on the cloud ERP and CRM vendors with their traditional counterparts and the new business is flowing one way. In a year's time, people will be looking at enterprise applications and the winners will be those who embraced cloud.