Study shows AI/ML is top workload for Kubernetes; edge IoT projects on the rise
There are two things that IT professionals seem to agree on: Kubernetes is a useful, widely used technology for managing containerized workloads. Also, it’s hard to use. Still, the trends revealed by a new study suggest that Kubernetes’ popularity will eventually spread to edge computing environments.
A study produced by Vanson Bourne for D2iQ, a provider of cloud platforms and services, found 75% of 300 organizations surveyed said they already use Kubernetes. Of those 40% are in production and another 35% are in the development or pre-production phase. Respondents were developers, DevOps employees (platform architects and cloud architects) senior IT decision-makers.
The top use of Kubernetes: AI and ML workloads, which 40% of respondents cited as their organization’s most popular workload. Indeed, 88% of respondents said expect Kubernetes will be the platform of choice for running AI and ML workloads in the next two years.
While developers appear to be mostly using centralized private and public cloud deployments at present time, AI and ML are one of the most frequently cited workloads the enterprises want to perform at the edge closest to their data source.
Kubernetes will overlap with edge computing in other ways. The report said that 81% of organizations are implementing “edge/IoT projects” on Kubernetes, with 61% running in production environments.
The news for developers and architects is not all goodness and growth — according to the survey, only 42% of organizations said all applications running on Kubernetes made it to “Day 2” environments, meaning that the applications are being maintained and optimized on an ongoing basis.
One of the challenges with Kubernetes is a lack of experience with the technology. On average, organizations that are using Kubernetes initially adopted it two years ago and expect to “complete their Kubernetes journey” in about two years, the report said.
As a reflection of the growing use and experience, the report noted the following:
- On average, 53% of all an organization’s projects are currently in production on Kubernetes, up from 42% in the 2020 survey.
- The average time that it took to get Kubernetes deployments to production was four and a half months, down a half month from 2020 results.
“This year’s report validates what we’ve been seeing as the Kubernetes market continues to mature,” said Tobi Knaup, founder and CEO, D2iQ. “Without the right technology and expertise in place, complexity challenges will kill Kubernetes deployments in Day 2 production environments.”
Article Topics
D2iQ | DevOps | edge AI | IoT | Kubernetes | market research
Comments