SigNoz

SigNoz is an open-source, OpenTelemetry-based observability platform for traces, metrics and logs, backed by ClickHouse. Self-hosted Community edition installs with Docker via Foundry; version 0.147.0.

OS: Linux / macOS (Docker), Kubernetes
Size: Docker images; 4 GB RAM for Docker
Version: 0.147.0

SigNoz: what it is

SigNoz is an open-source observability platform built on OpenTelemetry. It keeps traces, metrics and logs in one place, with dashboards, alerts and exception tracking on top, and stores the data in ClickHouse. Applications send data through standard OpenTelemetry SDKs and collectors, so you are not tied to a vendor-specific agent.

It is aimed at teams running services or microservices who want application performance monitoring and logs without a paid SaaS, or who need to keep telemetry on their own infrastructure. We have not deployed SigNoz ourselves (Not tested yet); the details below come from the SigNoz repository and documentation.

SigNoz at a glance

Item Details
Developer SigNoz Inc. (GitHub: SigNoz/signoz)
Latest version v0.147.0, released on GitHub on 22 October 2026
Editions Community (free, self-hosted), SigNoz Cloud, and Enterprise (Cloud, BYOC or self-hosted)
Licence Code outside the ee/ and cmd/enterprise/ directories is under the MIT (Expat) licence; enterprise code is under a separate licence in ee/LICENSE
Self-hosted platforms Docker on Linux or macOS (on Windows, through WSL 2 with Docker Engine inside WSL), Kubernetes, or Linux directly
Ports 8080 for the UI, 4317 (gRPC) and 4318 (HTTP) for OTLP ingestion
Last verified 27 October 2026, against the SigNoz GitHub repository and installation docs

What SigNoz covers

  • Distributed tracing: follow a request across services and see slow spans and errors.
  • Metrics and dashboards: application, infrastructure and custom metrics, queried with the Query Builder, PromQL or ClickHouse SQL.
  • Logs: search and filter logs and connect them to traces.
  • Infrastructure: Kubernetes clusters, pods, nodes and workloads, plus host CPU, memory, disk and network.
  • Alerts and exceptions: alert rules on your telemetry and a view of application exceptions.
  • AI features: an MCP server for bringing telemetry into coding agents. The Noz assistant is listed as available only on SigNoz Cloud.

How to self-host SigNoz with Docker

The old install.sh script and the Docker Compose files in the repository’s deploy folder are now deprecated. SigNoz installs through Foundry, an open-source CLI that describes the deployment in a config file. The documented Docker route:

  1. Check the prerequisites: Linux or macOS, Docker Engine 20.10 or later with the Docker Compose v2 plugin, at least 4 GB of memory for Docker, and ports 8080, 4317 and 4318 open.
  2. Install foundryctl with the script from signoz.io, or install it manually as described in the Foundry guide.
  3. Create a casting.yaml file that targets Docker (flavor: compose, mode: docker).
  4. Run foundryctl cast -f casting.yaml. It checks Docker, generates the Compose files and starts the containers. If you prefer to run Docker Compose yourself, use foundryctl gauge and foundryctl forge, then docker compose up -d in the generated folder.
  5. Check the containers with docker ps, open the UI on port 8080 and point your OpenTelemetry exporters at port 4317 or 4318.

If you already run SigNoz with the old Docker Compose files, the repository has a migration guide for moving to Foundry. On Windows, the documentation recommends Docker Engine inside WSL 2 rather than Docker Desktop, because of reported ClickHouse Keeper crashes under Docker Desktop.

SigNoz compared with Netdata and Zabbix

SigNoz is about applications: traces, logs and service metrics collected with OpenTelemetry. Netdata is a per-node agent with live, per-second system and service metrics. Zabbix is a central monitoring server for hosts and network devices, with templates and long-term history. They solve different problems, and many teams combine an infrastructure monitor with an APM tool.

Limitations to know

  • Self-hosting means running ClickHouse and several containers; plan memory (4 GB for Docker at minimum) and disk for your data volume.
  • Docker installs require Linux or macOS; on Windows only through WSL 2.
  • Some features are limited to the Cloud or Enterprise editions; check the edition comparison before relying on a feature.
  • The deployment method changed to Foundry. Older blog posts and tutorials with install.sh or docker-compose commands are out of date.

FAQ

Is SigNoz free?

The self-hosted Community edition is free. Most of the code is under the MIT licence, enterprise features are under a separate licence, and SigNoz Cloud and Enterprise are paid.

What port does the SigNoz UI use?

Port 8080 in the current Docker install. OTLP data is received on 4317 (gRPC) and 4318 (HTTP).

Can I run SigNoz on Windows?

Not natively. The documentation describes running it inside WSL 2 with Docker Engine installed in the WSL distribution.

Is SigNoz an alternative to Datadog or New Relic?

It covers the same core areas, namely traces, metrics, logs, dashboards and alerts, as an open-source, OpenTelemetry-native option you can host yourself. Feature parity depends on your needs, so compare specific features before switching.

What is the latest SigNoz version?

On 27 October 2026 the latest GitHub release was v0.147.0 (22 October 2026). SigNoz publishes new releases often.

Sources

Other articles

Send a Message