Skip to content

Apache SkyWalking

Apache Software Foundation APM and observability analysis platform for distributed systems, microservices, cloud-native, and service mesh environments.

Overview

Apache SkyWalking is a mature, enterprise-grade observability platform designed for distributed system, microservice, and cloud-native architectures. Governed by the Apache Software Foundation (ASF), it provides APM, distributed tracing, metrics, logging, event analysis, and topology visualization. Its architecture is highly pluggable, with BanyanDB as its native purpose-built observability database.

Repository & Community

Attribute Detail
Repository github.com/apache/skywalking
Stars ~24.8k ⭐
Latest Version APM 10.4.0 (April 1, 2026), BanyanDB 0.10.1 (April 6, 2026)
Language Java (OAP Server), Go (Satellite, BanyanDB)
License Apache 2.0
Governance Apache Software Foundation (ASF Top-Level Project)
Contributors ~996

Evaluation

  • Why it's better: Full ASF governance (no single-vendor risk), exceptional multi-language agent support (Java, .NET, Go, Python, Node.js, PHP, Rust, C++), native service mesh integration (Istio/Envoy), purpose-built BanyanDB reduces resource usage 5x vs Elasticsearch, and pioneering GenAI observability features.

  • When it fits (Applicability):

  • Java-heavy enterprise environments
  • Service mesh deployments (Istio/Envoy) needing topology analysis
  • Organizations requiring ASF governance and vendor-neutrality
  • Large-scale distributed systems (100B+ telemetry data points)

  • Teams needing multi-language auto-instrumentation agents

  • Pros and Cons:

Pros Cons
ASF governance (no vendor lock-in) Java ecosystem (JVM overhead for OAP)
Best-in-class Java agent Steeper learning curve than SigNoz/Coroot
10+ language agents BanyanDB still maturing (v0.10)
BanyanDB: 5x less RAM than ES Older UI compared to modern alternatives
Service mesh native (Istio/Envoy) Complex multi-repo ecosystem
GraalVM native image support Less OTel-native (proprietary agent format)
GenAI/LLM observability (v10.4) Helm chart complexity
eBPF network profiling Smaller cloud presence vs Grafana

Architecture

flowchart TB
    subgraph Probes["Probes & Agents"]
        JA["Java Agent"]
        NA[".NET / Go / Python<br/>Node.js / PHP Agents"]
        SW_eBPF["eBPF Agent<br/>(Rover)"]
        OTEL_P["OTel Collector"]
        Envoy["Envoy ALS<br/>(Service Mesh)"]
    end

    subgraph OAP["OAP Server<br/>(Observability Analysis Platform)"]
        direction TB
        Recv["Receiver Layer<br/>(gRPC, REST, Kafka)"]
        Core["Analysis Core<br/>(MAL, MQE, LAL)"]
        Alert["Alerting Engine"]
        Topo["Topology Analysis"]
    end

    subgraph StorageLayer["Storage (Pluggable)"]
        BDB["BanyanDB<br/>(recommended)"]
        ES["Elasticsearch<br/>/ OpenSearch"]
        CH_SW["ClickHouse"]
        PG["PostgreSQL"]
    end

    Probes --> OAP
    OAP --> StorageLayer
    OAP --> SWUI["SkyWalking UI"]

    style BDB fill:#1b5e20,color:#fff
    style OAP fill:#0d47a1,color:#fff

Key Components

Component Role
Language Agents Auto-instrument Java, .NET, Go, Python, Node.js, PHP, Rust, C++
OAP Server Core analysis engine: ingestion, aggregation, topology, alerting
BanyanDB Native observability database: 5x less RAM, 30% less disk vs ES
SkyWalking UI Web dashboard: topology, trace waterfall, metrics, logs
Satellite (Go) Sidecar/edge agent for proxy, rate limiting, load balancing
Rover (eBPF) Kernel-level network profiling agent

Sub-Projects & Versions (April 2026)

Sub-Project Version Status
SkyWalking APM 10.4.0 Production
BanyanDB 0.10.1 Production-ready
Java Agent 9.5.0 Stable
GraalVM Distro Experimental ~41 MiB idle vs ~1.2 GiB JVM
Satellite Latest Production
Rover (eBPF) Latest Production

Key Features

Feature Detail
Multi-Language APM Auto-instrumentation for 10+ languages
Service Mesh Native Istio/Envoy ALS topology analysis
BanyanDB Purpose-built columnar + time-series DB
eBPF Profiling Network profiling via Rover agent
GenAI Observability LLM latency, TTFT, token usage, cost (v10.4)
Topology Analysis Auto-discovered service dependency graphs
Alerting Rule-based + ML-powered anomaly detection
GraalVM Native Sub-ms startup, 30x less memory
100B+ Scale Proven at massive telemetry volumes

Compatibility

Dimension Support
Probe protocols Native SW, OpenTelemetry, Zipkin, Prometheus, Zabbix
Storage backends BanyanDB (recommended), Elasticsearch, OpenSearch, ClickHouse, PostgreSQL, H2
Service meshes Istio, Envoy (ALS), Linkerd
Languages Java, .NET, Go, Python, Node.js, PHP, Rust, C++, Lua
Platforms Kubernetes (Helm/Operator), Docker, bare metal, GraalVM native
CPU architecture amd64, arm64

Sources