25K ·GitHub
What is this?
Apache SkyWalking is an open-source APM (Application Performance Monitoring) system for monitoring, tracing and diagnosing distributed systems. It is built for microservices, cloud native and container-based architectures.
Apache SkyWalking is an open-source APM (Application Performance Monitoring) system for monitoring, tracing and diagnosing distributed systems. It is built for microservices, cloud native and container-based architectures.So is it a server I deploy, or something I plug into my applications?Both. You deploy a SkyWalking cluster that collects and analyzes telemetry and add agents to your apps; a separate eBPF agent, Rover, profiles Kubernetes deployments for CPU and network performance.
Both. You deploy a SkyWalking cluster that collects and analyzes telemetry and add agents to your apps; a separate eBPF agent, Rover, profiles Kubernetes deployments for CPU and network performance.How do I get it running?Download a release from the releases page, or compile the project yourself with the documented build steps. The website also offers a live demo and a showcase for standing up a preview deployment quickly.
Download a release from the releases page, or compile the project yourself with the documented build steps. The website also offers a live demo and a showcase for standing up a preview deployment quickly.Which languages can it instrument?There are agents for Java, .Net Core, PHP, NodeJS, Golang, Lua, Rust, C++, client-side JavaScript and Python, all in active development and maintenance.
There are agents for Java, .Net Core, PHP, NodeJS, Golang, Lua, Rust, C++, client-side JavaScript and Python, all in active development and maintenance.Is there a ceiling on how much telemetry it can handle?Not per the README: it claims 100+ billion telemetry data points can be collected and analyzed from a single SkyWalking cluster.
Not per the README: it claims 100+ billion telemetry data points can be collected and analyzed from a single SkyWalking cluster.Is it free to use, including commercially?Yes, it is Apache 2.0 licensed, and hundreds of companies and organizations use it in production and for commercial purposes.
Yes, it is Apache 2.0 licensed, and hundreds of companies and organizations use it in production and for commercial purposes.We already run Prometheus, Zipkin or OpenTelemetry — do we have to switch?No. SkyWalking ingests metrics, traces and logs from mature ecosystems like Zipkin, OpenTelemetry, Prometheus, Zabbix and Fluentd, and processes known meter formats through the same script pipeline as its native format.
No. SkyWalking ingests metrics, traces and logs from mature ecosystems like Zipkin, OpenTelemetry, Prometheus, Zabbix and Fluentd, and processes known meter formats through the same script pipeline as its native format.Where does all the collected telemetry get stored?In BanyanDB, SkyWalking's own observability database created in 2022 to ingest, analyze and store telemetry data.
In BanyanDB, SkyWalking's own observability database created in 2022 to ingest, analyze and store telemetry data.Does it do anything smarter than dashboards and alerts?Yes — ML and AI analyze the observability data: they recognize HTTP URI patterns, calculate metric baselines for intelligent alerting and improve anomaly detection. You can also set service-, deployment- or API-centric alarm rules and forward alerts to third-party tools.
Yes — ML and AI analyze the observability data: they recognize HTTP URI patterns, calculate metric baselines for intelligent alerting and improve anomaly detection. You can also set service-, deployment- or API-centric alarm rules and forward alerts to third-party tools.Can I try it before deploying anything?Yes: there is a live demo with the native UI and Grafana, an interactive 3D infrastructure map, and a showcase guide for setting up a preview deployment.
Yes: there is a live demo with the native UI and Grafana, an interactive 3D infrastructure map, and a showcase guide for setting up a preview deployment.What are similar projects?openobserve, signoz, VictoriaMetrics, pinpoint, cat, kubeshark, evenfire, loki
openobserve, signoz, VictoriaMetrics, pinpoint, cat, kubeshark, evenfire, lokiWhich topics is it in?performance-monitoring, distributed-tracing, service-mesh, web-performance, apm, ebpf, sre, grafana, metrics, logging, prometheus, cloud-native
performance-monitoring, distributed-tracing, service-mesh, web-performance, apm, ebpf, sre, grafana, metrics, logging, prometheus, cloud-native