When it comes to cloud computing trends, little remains the same from one year to the next. In 2018, we witnessed a year filled with explosive growth of the cloud, big data analytics and SaaS-based business applications and services. And as businesses modernize their computing and networking architectures, cloud-native architectures have remained one of the hottest technology trends.
The reason? The way enterprises do business is changing. With the emergence of microservice based architectures, companies are able to bring innovative products to market faster and more cost effectively than ever before. Cloud services have evolved from back-end services used to support the business to critical components of enterprise strategy to create differentiation and drive the business forward.
In 2019, I can say with certainty that multi-cloud is here to stay. And as the proliferation of multi-cloud continues over the next twelve months, it will be imperative for enterprise IT teams to continue to adapt to the new realities of modern application architectures in order to remain competitive. Here are a few of my thoughts surrounding the main trends that will drive the multi-cloud industry in 2019.
Uptick in Kubernetes
Last year, we saw upstream Kubernetes reach sufficient maturity. As such, vendors are no longer required to add custom code, or modify the offering as we saw in the past with Rancher and OpenShift. When you combine this maturity with the managed Kubernetes offerings from the hyperscale public clouds, it’s one step closer to the promise of application portability. However, there are a few exceptions such as RBAC, which is backed by completely different identity providers on each cloud, and storage, which still requires you to step outside Kubernetes to migrate, replicate and backup data.
Installing Kubernetes will remain a challenge, though with the above-mentioned managed offerings, plus Platform as a Service (PaaS) offering from Red Hat and Pivotal, this ceases to be an issue for many use cases. Production grade installs via Kubeadm will remain a work in progress for the time being, but they should be here any minute now.
Lastly, we can expect visibility and governance of Kubernetes deployments to remain an issue for organizations using traditional IT and change management tools. This is because organizations need to rethink and adapt their processes for use with Kubernetes, or they won’t be able to keep up with the rapid and automatic changes that will occur.
DevOps and Security Collide
The way enterprises use the cloud is changing. Next-generation applications, built and delivered using DevOps methodology, are moving to the cloud. These virtual environments continue to evolve as enterprises move their IT to the cloud, enabling the scalability and flexibility of on-demand resources. But with next-generation applications come the need for increased security, as many of these applications are not only complex, but can potentially have different types of security issues ranging from misconfigured servers to vulnerable code.
Fortunately, we saw security vendors really step up their game in 2018. Container registries such as Harbor, JFrog and Google Container Registry added vulnerability scanning, better capabilities to restrict access to risky containers, and track changes with content signing. In 2019, it will be up to DevOps teams to make use of these capabilities to bring security teams in to the fold with DevSecOps, as more organizations realize traditional security practices can’t scale.
Machine Learning Will Remain A Key Component of Automation
Automation in the cloud has proven itself a critical capability for organizations in their ability to manage every cloud asset across a network – from expansive private cloud infrastructures to single, public cloud instances spun up by an individual developer (i.e. shadow IT). This complex environment will need more than just automation to continue to run smoothly. Machine learning algorithms will play a big part in the latter half of 2019, especially when it comes to remediation and predictions of multi-cloud failures.
With the majority of organizations deploying at least two cloud solutions, multi-cloud is here to stay for the next few years. However, in order to get the most out of multi-cloud strategies in 2019, organizations need to take advantage of cloud native technologies, embrace DevSecOps, and not be afraid to incorporate machine learning into automation strategies.
