# microsoft/retina

eBPF distributed networking observability tool for Kubernetes

Repository: https://github.com/microsoft/retina
Canonical: https://ross.abutalabs.com/products/retina
Homepage: https://retina.sh
Language: Go
License: MIT
License Family: permissive
Topics: ebpf, kubernetes, networking, observability
Last push: 2026-08-26T17:24:24+00:00

## Health v2 (maintenance only)
Score: 92/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 98, longevity 68
- inputs: {"age_days": 954, "days_push": 7, "days_rel": 12, "gap_med": 12.0, "n_releases_24m": 37}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3166, forks 303 (observed 2026-08-28T04:07:46.708818+00:00)

## What it is
Retina is an open-source, eBPF-based Kubernetes network observability platform that collects customizable telemetry, Prometheus metrics, and distributed packet captures. It is cloud- and CNI-agnostic, exporting data to Prometheus, Azure Monitor, Grafana, and other backends.

## Use cases
- debug pod-to-pod network connectivity issues in kubernetes
- monitor cluster network health with prometheus metrics
- run distributed packet captures across nodes for troubleshooting
- get network flow logs and DNS insights via hubble integration
- monitor network security and compliance in kubernetes clusters
- troubleshoot why pods cannot connect to each other

## When to choose
- you need eBPF-based network observability for Kubernetes across Linux and Windows nodes
- you want cloud- and CNI-agnostic monitoring that works on any Kubernetes distribution
- you need on-demand distributed packet captures without manually running tcpdump on nodes
- you want Prometheus metrics exported to Grafana or Azure Monitor

## When to avoid
- you run high-core-count nodes (32+ cores) under heavy load with advanced metrics, given reported performance limitations
- you still rely on Windows Server 2019 nodes, which are no longer supported
- you need a general-purpose host-level eBPF observability tool outside Kubernetes

## Facets
- artifact type: application
- maturity: active
- function: monitoring, tracing, networking, cli
- domain: networking, monitoring
- platform: windows, go, self-hosted, cloud
- tags: ebpf, network-observability, packet-capture, prometheus-metrics, hubble, cni-agnostic, helm, kubernetes, devops, containers, linux, docker

## Member repositories
- microsoft/retina (main) score 92

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:46.708818+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:25:10.246297+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/retina (fetched 2026-08-28T04:07:46.708818+00:00, sha 55ad5aedb51c)
  - homepage: https://retina.sh (fetched 2026-08-29T09:39:46.306483+00:00, sha 3630eca107bc)
  - site_page: https://retina.sh/docs/Introduction/intro (fetched 2026-08-29T09:39:46.315773+00:00, sha 1e5d562721e4)
- Data as of 2026-08-30T08:39:29.467469+00:00.