The Microsoft Azure Container Networking squad is excited to denote Retina, a cloud-native instrumentality networking observability level that enables Kubernetes users, admins, and developers to visualize, observe, debug, and analyse Kubernetes’ workload postulation irrespective of Container Network Interface (CNI), operating strategy (OS), and cloud. We are excited to merchandise Retina arsenic an open-source repository that helps with DevOps and SecOps related networking cases for your Kubernetes clusters and we invitation the open-source assemblage to innovate on with us.

Embracing and advancing open-source software

Cloud autochthonal technologies similar Kubernetes person made gathering applications that tin tally anywhere, easier. At the aforesaid time, galore applications person go much complex, and managing them successful the unreality is progressively difficult. As companies physique cloud-native applications composed of interconnected services and past deploy them to aggregate nationalist clouds arsenic good arsenic their backstage infrastructure, web related observability, troubleshooting, and debugging has go progressively difficult.

With the powerfulness of extended Berkley Packet Filter (eBPF), it is present imaginable to connection actionable web insights including however containerized micro-services interact and bash truthful successful non-intrusive ways without immoderate alteration successful the applications itself—that’s precisely what Retina sets retired to achieve. Retina volition assistance democratize web observability and troubleshooting by bringing caller absorption to the acquisition of exertion developers. Retina provides developers with elemental ways to observe and troubleshoot their applications for issues specified arsenic packet drops and latency without worrying astir the complexities of the underlying web infrastructure and transformations.

Based connected our affirmative acquisition successful the assemblage with eBPF and Cilium, we are excited to physique connected this narration and prosecute some much intimately and with much communities. We judge that by opening Retina to the community, we tin payment from informed feedback, innovative ideas, and collaborative efforts that volition assistance heighten and grow Retina’s capabilities.

Retina solutions and capabilities

Drawing from our extended acquisition managing aggregate instrumentality networking services for the Azure Kubernetes Service (AKS), we identified captious gaps successful web monitoring, the postulation of web metrics and traces from Kubernetes clusters. Retina is simply a cutting-edge solution that closes these gaps and is designed to tackle the analyzable challenges of managing and supporting Kubernetes networks providing infrastructure- and site-reliability engineers broad insights into clump networking. Retina besides provides heavy postulation investigation with Kubernetes-specific context, translating metrics into either industry-standard Prometheus oregon web travel logs.

Existing open-source solutions are often tightly coupled with circumstantial CNI’s, OS, oregon information planes, thereby limiting their versatility and use. For this reason, Retina has been designed and developed to beryllium a highly versatile, adaptable, and extensible model of plugins susceptible of moving seamlessly with immoderate CNI, OS, oregon unreality provider—making it a invaluable summation to immoderate existing toolset. Retina supports some Linux and Windows information planes, ensuring it meets the divers needs of infrastructure- and site-reliability engineers, portion maintaining a minimal representation and CPU footprint connected the cluster—this remains existent adjacent astatine scale. Retina’s pluggability plan ethos helps america easy widen and accommodate to code caller usage cases without depending connected immoderate circumstantial CNI, OS, oregon information plane.

Figure 1: Architecture overview of Retina

One of Retina’s cardinal features provides heavy web postulation insights that see Layer 4 (L4) metrics, Domain Name System (DNS) metrics, and distributed packet captures. It seamlessly integrates the Kubernetes app exemplary offering pod-level metrics with elaborate context. It emits actionable networking observability information into industry-standard Prometheus metrics providing node-level metrics (for example, forward, drop, Transmission Control Protocol (TCP), User Datagram Protocol (UDP), and Linux utility) and pod-level metrics (such arsenic basal metrics, DNS, and API server latency.)

Retina’s distributed packet captures are label-driven—allowing users to specify what, where, and who to seizure packets from. Additionally, it provides humanities discourse of web travel logs and precocious debugging capabilities that heighten web troubleshooting and show optimization.

Our imaginativeness for Retina

Many enterprises are multi-cloud and privation solutions that enactment good not conscionable connected Microsoft Azure, but connected different clouds arsenic good arsenic on-premises. Retina is open-source and multi-cloud from time one. By open-sourcing Retina, we purpose to stock our cognition and imaginativeness for Kubernetes networking observability with the broader cloud-native ecosystem. Our anticipation is that Retina volition germinate and turn done collaboration with different developers and organizations who stock akin experiences and goals successful this field.

In presumption of architecture, extensibility was cardinal from the outset and volition stay going forward. Retina offers extensibility successful information collection—allowing users to easy adhd caller metrics and insights. It besides offers extensibility successful exporters—enabling users to integrate with different monitoring systems and tools. This flexibility ensures that Retina tin accommodate to antithetic usage cases and environments, making it a versatile and almighty level for Kubernetes networking observability. In conclusion, we envision Retina arsenic a level allowing anyone to contribute, extend, and innovate connected yet creating a robust, purpose-built, and broad solution for Kubernetes networking observability.

Getting started

We anticipation this blog station has fixed you an overview of Retina, its features, and benefits. Retina is simply a task that we are precise excited about, and we invited your feedback, suggestions, and contributions. Whether you are a web engineer, a DevOps engineer, a tract reliability engineer, oregon a cloud-native enthusiast, we invitation you to effort retired Retina and spot however it tin assistance you summation deeper visibility into your Kubernetes networks.

• eBPF distributed networking observability tool for Kubernetes
• Introduction to Retina
• What is Azure Kubernetes Service (AKS) Network Observability?
• If you person questions, reach retired to us.

The station Microsoft unfastened sources Retina: A cloud-native instrumentality networking observability platform appeared archetypal connected Microsoft Azure Blog.