Survey Highlights Kubernetes Ascendancy in the Enterprise
A survey of 527 IT professionals published today finds that 80% plan to build and deploy the bulk of new applications are being deployed on Kubernetes clusters, with well over half (58%) also planning to migrate legacy workloads to the cloud0natove platform as well.
Conducted by Dimensional Research on behalf of Pure Storage, the survey also finds that among those migrating workloads, nearly two-thirds plan to do so within the next two years. A full 85% said they plan to move most of the workloads they have running on virtual machines to the cloud-native platform, with 71% noting a unified platform for centralizing the management of virtual machines and containers would greatly benefit their organization.
The recent acquisition of VMware by Broadcom appears to be a significant factor in that decision. A total of 80% said the acquisition of VMware made them nervous about the future of their existing virtual machine-based workloads. A total of 79% say the primary drivers for migration are scalability, flexibility, operational simplicity and cost considerations.
Murli Thirumale, general manager for the Cloud Native Business unit for Pure Storage, said organizations’ biggest challenge when making that migration is the different management paradigm. There is no way to, for example, duplicate the VMware vCenter management platform, so IT teams need to be retrained to manage Kubernetes environments.
Overall, the survey finds that organizations are now deploying a wider range of stateful applications on Kubernetes clusters, with nearly all (98%) of respondents run data-intensive workloads such as databases (72%), analytics (67%), and artificial intelligence (AI) and machine learning (ML) workloads (54%). Major benefits of deploying these applications on Kubernetes clusters have been faster deployments (65%), followed by reduced infrastructure costs (54%) and increased uptime (53%).
According to the survey, most of those applications are also being deployed across hybrid cloud environments (86%). However, challenges remain, with more than half (52%) citing higher availability and disaster recovery as a challenge and 46% also desiring faster storage systems.
Most survey respondents (91%) have at least two years of Kubernetes experience, with 56% having more than four years. A full 96% said they already have platform engineering teams in place to increase the scalability and flexibility of their applications. Organizations have invested in training (63%), consultants (60%) and hiring skilled engineers (52%) to support this function. A large majority (82%) said that platform engineering is a function, not a job title.
In the longer term, the overall platform engineering goal should be to make Kubernetes clusters all but invisible to developers of cloud-native applications by providing more self-service capabilities, said Thirumale. Today, developers are still required to know too much about the underlying infrastructure to succeed, which in turn limits the number of applications being built and deployed, he added.
It’s not clear when the bulk of most applications might one day be deployed on Kubernetes clusters but it’s clear more organizations are starting to view Kubernetes as a de facto standard. The challenge now is finding enough expertise to manage what are rapidly becoming fleets of Kubernetes clusters distributed across the enterprise that with advances in AI will soon become much simpler to deploy, secure and update.
