Grafana Labs Allies With Isovalent to Improve Cloud-Native Observability
Grafana Labs and Isovalent have integrated open source Grafana dashboards with the Cilium network overlay for containers that runs at the extended Berkeley Packet Filter (eBPF) microkernel level of the Linux operating system.
Isovalent CEO Dan Wendlandt says the integration will enable IT teams to employ Prometheus metrics, Grafana dashboards, alerts and Grafana Tempo traces to observe the entire cloud-native computing stack from Layers 2 through 7 without having to employ container sidecars to instrument applications.
As a result, observability is going to be one of the killer use cases for eBPF, he says. For the first time, it will become possible to leverage the visibility that Cilium provides to attach identities of applications and end users to a suite of observability tools instead of continuing to rely on IP addresses, notes Wendlandt.
As a kernel-level technology, eBPF allows software to run in a lightweight, sandboxed environment without needing to add additional modules or modify the kernel source code. That approach enables services ranging from networking and storage to monitoring and security to be more efficiently delivered because they no longer need to run in user space alongside other applications. In effect, it bridges the boundary between kernel and user space by enabling developers to combine and apply logic across multiple subsystems that, historically, were completely independent of one another.
Cilium, meanwhile, is being advanced as an incubation-level project under the auspices of the Cloud Native Computing Foundation (CNCF). Originally created by Isovalent, Cilium is at the forefront of a wave of networking, security, storage and monitoring platforms that are expected to employ eBPF to dramatically improve performance at scale. The CNCF is an arm of the Linux Foundation, which earlier this year also launched the eBPF Foundation that is committed to expanding the use of eBPF across all operating systems.
IT organizations would be well-advised to ask their vendors when they plan to support eBPF. Cloud service providers generally require it to deliver their own managed services more efficiently. The issue then becomes determining which platforms will make use of eBPF sooner than later; otherwise, IT teams might be looking at an upgrade sooner than they might normally expect. It is already apparent upgrades are in the offing for not just networking, storage and security platforms but also instances of operating systems that don’t yet support eBPF.
At the same time, networking is becoming a bigger challenge as the number of Kubernetes clusters being deployed in production environments continues to expand. IT teams are being increasingly tasked with networking multiple Kubernetes clusters together across what is becoming a highly distributed computing environment. Cilium is already gaining traction as an open source network overlay that already supports eBPF at the expense of network overlays that require IT teams to install some type of virtual machine software.
It’s not quite clear to what degree networking in cloud-native application environments will require dedicated specialists versus becoming more of an extension of a DevOps workflow. Regardless of approach, however, networking is about to play a more pivotal role that goes well beyond the need for simple connectivity.
