How WebAssembly can help CSPs look beyond Kubernetes for faster, greener, and more secure software deployment

Commercial context

CSPs working on software deployments are increasingly asking “is there life after Kubernetes?” Most work in this area today focuses on virtual machines (VMs) and containers – yet there is an increasingly clear need to raise expectations for efficient, portable and secure application hosting. Docker containers, the preferred option for now, are full Linux stacks which take several seconds to start – the future however lies in cloud native software which is portable across environments while also meeting new standards for speed, security and sustainability.

This is where WebAssembly (Wasm) – a W3C standard for universal computing which is smaller, faster, cheaper, safer and more portable than Docker containers – comes in. Just as the industry eagerly embraced containers and Kubernetes over machine virtualization, the emergence of Wasm promises another leap forward with lightweight, instantaneous processes, presenting a significant evolutionary step beyond containers. Wasm serves as a stack-based virtual machine for clients and servers, acting as a portable compilation target for high-level languages – offering major cost and performance benefits to use cases across ODA Canvas, edge computing and core networks.

The solution

The WebAssembly Canvas Catalyst is therefore demonstrating the benefits of this emerging technology for CSPs. This has been done by demonstrating an ODA Canvas realized on WasmCloud (a CNCF sandbox project for simple distributed application development and hosting using Wasm actors and capability providers) to compose ODA components. Open-source implementations of ODA components and other functions can be demonstrated in a running cluster of low-power ARM servers, proving the advantages of using Wasm runtime environments.

This approach enables CSPs to distribute ODA and deploy software anywhere – from cloud, to edge, to customer devices – saving energy and integration costs, and improving security and agility at the same time. The Catalyst therefore demonstrates Wasm’s potential for rapid development, ease of multi-vendor integration, and platform portability. This approach to universal compute enables enormous latitude in deployment options, with marked carbon footprint reduction across the IT infrastructure spectrum. Architecture-independent packaging greatly simplifies operations across heterogenous environments – in the core, at the edge and in user equipment – and the instantaneous process instantiation and sandboxing inherent in Wasm enables scalable commercial product offerings based on functions-as-a-service (FaaS) models.

Wider value

More broadly, this project points to a new approach to software development and microservices – opening up a wide range of improvements in how applications work together. The ability to code without burdensome boilerplate scaffolding offers runtimes which are a fraction of the current norm. Wasm is essentially a universal language which is not dependent on server architecture, and can be deployed on a vast range of devices (including end-user devices) - helping programs run more efficiently, without the need for hardware-specific compilations, while the devices involved can achieve better results while also using less energy and producing less heat. While still an emerging technology Wasm offers potentially enormous benefits to the telecommunications industry in terms of speed, reliability, and sustainability.