Skip to content
OpenSourceSoftware.ai
SigNoz logo

SigNoz

SigNoz · Observability

GitHub stars

32,492
—Awaiting first sync

Awaiting first sync

No measured star history yet

What is SigNoz?

SigNoz keeps logs, metrics, traces, dashboards and alerts in one application, so you can move from a slow service chart to the exact trace and its logs without switching tools. It is built on OpenTelemetry, which means the instrumentation in your code stays standard if you ever change backends.

SigNoz is an observability platform built on OpenTelemetry. It collects logs, metrics and traces from your applications and servers, stores them in ClickHouse, and shows them together in dashboards, alerts, trace views and exception lists, so a team can follow one problem across every signal.

Core capabilities

  • Search logs, follow traces and chart metrics side by side, linked to the same request.
  • Build dashboards with a visual query builder, PromQL or ClickHouse SQL, and alert on any query.
  • Send data with OpenTelemetry and Prometheus, so instrumentation is not tied to one vendor.

Logs, metrics and traces together

The APM view shows latency, error rate, throughput and Apdex for each service, with top endpoints, database calls and external calls. Distributed tracing adds flamegraphs and waterfalls, and log management lets you search, aggregate and link logs to traces. Dashboards accept a visual query builder, PromQL or ClickHouse SQL, and the same queries can drive alerts. Infrastructure monitoring covers hosts and Kubernetes, and LLM apps can be traced for tokens, latency and cost.

  • APM
  • Tracing
  • Logs
  • Dashboards
  • Alerts

Query telemetry from coding agents

An optional MCP server, which self-hosted users run alongside their install, lets AI clients query metrics, logs and traces in plain language, review and edit alert rules, manage dashboards and open trace details. The in-app assistant called Noz is a separate feature that runs only on SigNoz Cloud.

  • MCP
  • AI agents

Self-hosting and licence

The free Community edition runs on Docker with the Foundry CLI, on Kubernetes with Helm, or on Linux. The Docker setup starts ClickHouse, PostgreSQL, the SigNoz OpenTelemetry collector and the SigNoz server, needs at least 4GB of RAM for Docker and serves the UI on port 8080. Code outside the ee/ and cmd/enterprise/ folders is MIT licensed. Anomaly detection, SAML, audit logs, fine-grained roles and multiple ingestion keys belong to the paid SigNoz Cloud and Enterprise plans.

  • Docker
  • Kubernetes
  • ClickHouse
  • MIT
  • Open core

Where it fits

Use cases

  1. 01

    Find why an API got slow

    Open the service in the APM view to see latency, error rate and top endpoints, then click into a slow trace. The flamegraph shows which database or external call took the time, with the related logs alongside.

  2. 02

    Replace a paid logging bill

    Point your OpenTelemetry collectors or agents at your own SigNoz server and ingest application logs there. Search and filter them with the query builder and save the views your team uses every day.

  3. 03

    Watch a Kubernetes cluster

    Install SigNoz with Helm and turn on the Kubernetes integration to collect node, pod and workload metrics. Build a dashboard for CPU, memory and disk, and set alerts that notify your on-call channel.