Cosmonic Open Source Project Integrates Nomad and Cilium
Cosmonic has launched an open source Netreap project to integrate Nomad, an alternative for orchestrating containers from Hashicorp, with the Cilium networking overlay for cloud-native applications created by Isovalent.
Dan Norris, head of infrastructure for Cosmonic, said the company uses both Nomad and Cilium within its platform-as-a-service (PaaS) environment for building and deploying applications based on WebAssembly (Wasm). Wasm is a portable binary instruction format for building software that runs in a memory-safe, sandboxed execution environment.
Netreap extends the Kubernetes container network interface (CNI) to now integrate Nomad with Cilium, which is now being advanced under the auspices of the Cloud Native Computing Foundation (CNCF) alongside Kubernetes. Cilium previously only worked across Kubernetes clusters, but Cosmonic has now extended its reach to Nomad.
The challenge many IT organizations are encountering with Kubernetes is that managing the platform can be a full-time job. Nomad provides a lighter-weight alternative for orchestrating containers that is simpler to manage.
It’s unclear whether Nomad will gain traction as an alternative to Kubernetes, but as the number of Nomad nodes deployed in cloud-native IT environments continues to increase, the need to network them becomes a priority for organizations that have adopted Nomad. Some organizations are also running a mix of Kubernetes and Nomad nodes that may need to share access to the same network.
Cilium already takes advantage of the extended Berkley Packet Filter (eBPF) capability that makes it possible to process networking, storage and security tasks in a sandboxed environment that runs in the Linux kernel rather than in user space. As a result, the overall performance of Cilium is much faster than legacy approaches to creating virtual networks.
That level of performance is required to drive networking in cloud-native application environments where microservices running on disparate pods within multiple clusters need to be integrated. The rise of application networking enabled by proxy software and service meshes running across networks such as Cilium.
The implications of application networking are profound in that it presents an opportunity to programmatically integrate network and security operations into DevOps workflows. Rather than requiring developers to master lower-level networking APIs, application networking provides a higher level of abstraction for invoking those services without having to rely on a network specialist.
IT organizations may soon need to revisit how they are structured. There may always be a need for dedicated networking specialists to manage the physical network underlay, but the networking services themselves will inevitably become more integrated with other approaches to managing infrastructure-as-code (IaC) as network operations increasingly becomes an extension of a DevOps workflow.
Regardless of how application networking evolves, it’s clear the rigidity that has characterized the delivery of network services for decades is fading as application environments become more distributed. The challenge is determining the path of least resistance for delivering the network services upon which those latency-sensitive application services depend.
