AT&T and Dell Technologies Team Up to Advance Container Platforms and Bare-Metal Kubernetes

Source: containerjournal.com (original reporting, August 2019)

Overview

AT&T and Dell Technologies announced a collaboration to advance several open source projects focused on container orchestration and cloud-native infrastructure. The partnership centers on reducing friction in deploying and managing containerized workloads at scale — both on virtual machines and, increasingly, on bare-metal servers.

Key Open Source Projects in Scope

The collaboration targets three specific open source initiatives:

• AirShip — A management framework that combines OpenStack and Kubernetes to provide declarative, automated cloud provisioning and lifecycle management for containerized applications. AT&T originally launched the project, and both companies committed to accelerating its next major release.
• Metal3-io — A project focused on simplifying the provisioning of Kubernetes clusters directly on bare-metal hardware, removing the dependency on a hypervisor layer.
• OpenStack Ironic — The OpenStack service for bare-metal provisioning. Both companies committed contributions to improve the process of deploying OpenStack directly on physical servers.

Both companies also pledged contributions to the Kubernetes Cluster API, which provides a declarative API for cluster lifecycle management and is central to integrating Kubernetes with infrastructure providers such as OpenStack and bare-metal environments.

The Case for Bare-Metal Kubernetes

Eric Vallone, Director of Service Provider Solutions at Dell Technologies, stated that Dell is committed to enabling joint deployments of OpenStack and Kubernetes — whether on virtual machines or bare-metal servers. At the time of the announcement, virtual machine-based deployments were the dominant pattern for both platforms.

However, running Kubernetes directly on bare-metal offers meaningful advantages for certain workloads:

• Reduced overhead — Eliminating the hypervisor layer removes the performance and resource cost associated with running workloads inside virtual machines.
• Heterogeneous hardware support — Bare-metal Kubernetes clusters can more readily incorporate specialized processors such as GPUs and FPGAs, which are increasingly important for AI, ML, and network processing workloads.
• Tighter hardware utilization — Containerized workloads on bare metal can take full advantage of physical server resources without the abstraction tax of a guest OS and hypervisor.

Vallone acknowledged that this trajectory has implications for VMware — a Dell Technologies subsidiary — whose core business is built around hypervisor-based virtualization. Dell’s apparent bet is that VMware can evolve its portfolio to serve as a management layer bridging legacy virtualized environments and modern cloud-native, Kubernetes-based infrastructure.

AT&T’s 5G Containerization Strategy

AT&T’s motivation for the collaboration is rooted in its 5G network transformation. The carrier has publicly committed to replacing proprietary network infrastructure with open source software running on industry-standard hardware. Containers are central to that strategy: deploying 5G network functions as containerized workloads enables AT&T to build, update, and scale services with the agility that software-defined networking demands.

Beyond internal network functions, AT&T has signaled an intent to cultivate a third-party developer ecosystem — allowing external developers to build and deploy containerized applications on AT&T-managed platforms.

To support this infrastructure shift, AT&T will deploy a mix of Dell-branded and white-box servers in place of the proprietary infrastructure used for 3G and 4G network delivery, according to Vallone. This positions Dell to build expertise in carrier-grade, cloud-native deployments that it can then translate into commercial offerings for enterprise customers undergoing similar transformations.

Broader Industry Implications

The AT&T–Dell partnership reflects a wider industry trend: major carriers and infrastructure vendors are redirecting significant engineering resources toward cloud-native computing frameworks. As both AT&T and Dell face competitive pressure in their respective markets, their investment in open source container infrastructure is likely to accelerate adoption across the broader telco and enterprise IT sectors.

Projects like AirShip, Metal3-io, and the Kubernetes Cluster API benefit directly from this kind of large-scale industrial participation — contributions from organizations operating at carrier scale tend to surface integration requirements and edge cases that smaller contributors may not encounter.

Summary