- 지나친 깨끗함은 더러움만 못하다?!··· AI 위한 데이터 관리 체크포인트
- 7 modi per creare il proprio Large Language Model
- Hulu's Black Friday deal is back: Get 1 year of Hulu for $0.99 a month, but sign up soon
- My favorite ultraportable laptop for work is not a ThinkPad or MacBook (and it's $350 off)
- The 2-in-1 laptop I recommend to most people is not a Dell or Lenovo (and it's $360 off)
Red Hat buttresses edge features in RHEL 8.4
New features in Red Hat Enterprise Linux (RHEL) are tuned to provide better remote support for edge networking where processor- and memory-constrained devices can present management problems.
RHEL 8.4 announced this week at Red Hat Summit has new capability to send lighter-weight universal base images and is designed for potentially less capable edge devices, letting Red Hat customers deploy edge applications more flexibly.
In addition to the new RHEL version, Red Hat announced updates to Podman, the company’s open-source container engine, that will allow users to manage widely deployed containers from a single console, and an OpenShift update that adds support for smaller clusters and remote worker nodes makes it easier to use Kubernetes in resource-constrained locations.
This offers a lot of upside to companies in the process of deploying edge computing systems, according to Dave McCarthy, vice president of IDC’s infrastructure practice. The new features offer more granularity in deploying over-the-air updates to edge devices that can be computationally limited or have restricted access to broadband networks. As a result, the ability to push only parts of a software image out as an update can be much more efficient than rolling out an entire new one, for instance.
“So when you have a device in an edge deployment, the last thing you want to do is brick it and have to go and update it by hand,” he said. “So now you have intelligent rollback, remote updating, scheduling—lots of things that limit downtime.”
According to Christian Renaud, research director at 451 Research, Red Hat’s moves are in lockstep with the development of the edge-computing market, which is maturing quickly as deployments move past proof of concept and into larger scale deployments of edge devices. “The edge market has gone from ones and twos to hundreds and thousands of devices being deployed,” he said.
That change is happening quickly, and more advanced features are in demand in a range of settings. In industrial facilities, edge computing is used for things like predictive maintenance and process optimization by running real-time analysis close to the factory floor. Retailers are increasingly interested edge computing to integrate inventory management, point-of-sale, and employee management into a discrete, remotely manageable framework.
Red Hat’s efforts aim at offering a flexible software base that can serve in diverse environments. “Edge is a fairly horizontal market,” said Cliff Grossner, vice president for research at the Dell’Oro Group. “We all think of Google [and] Amazon as having these centralized data centers, and they do, but they also have hundreds of points of presence across the globe, to be close to end users,” he said.
And while Google and Amazon have dominated mind-share around edge computing, their offerings represent vertically integrated, proprietary solutions that Red Hat has built a business out of challenging, said Renaud. “We’re at the beginning of the edge market, in its awkward, lanky teenage years, so it’s not a foregone conclusion that people are going to stay with those vendors,” he said. “There are always enterprises who want to hedge their single-vendor risk by adopting open source solutions.”
Red Hat also announced that the company is working on a certified Linux platform for road vehicles. Partnering with functional-safety and regulatory expert Exida, the company will work to develop an open-source software base for carmakers to use in their vehicles, Red Hat said.
Copyright © 2021 IDG Communications, Inc.