SigNoz
OpenTelemetry-native observability platform for logs, metrics, traces, alerts, and exceptions — a self-hostable Datadog and New Relic alternative.
SigNoz is an open-source, OpenTelemetry-native observability platform. It puts logs, metrics, traces, exceptions, alerts, and dashboards in one query surface, so you can move from a spiking service chart straight into the spans, log lines, and infrastructure metrics behind it. The community edition is AGPL-3.0 and runs entirely in your own infrastructure.
On top of the APM overview — latency, error rate, throughput, Apdex, top endpoints, database calls, and external calls — SigNoz ships distributed tracing with flamegraph and waterfall views, trace funnels for request-flow drop-offs, a visual query builder that also accepts PromQL and ClickHouse SQL, Kubernetes and host infrastructure monitoring, and LLM/AI observability for prompts, tool calls, token usage, latency, and cost. Alerting covers metrics, logs, traces, anomalies, and exceptions, with routing policies, planned maintenance windows, a Terraform provider, and a Kubernetes operator. The SigNoz MCP server exposes your telemetry to AI coding agents, and traces, metrics, and logs land in ClickHouse while Postgres handles metadata.
Compared with Datadog, New Relic, and Sentry, SigNoz avoids per-host, per-seat, and custom-metric pricing, and self-hosting keeps the whole data plane on your own machines. Install with the Foundry CLI against Docker, Kubernetes, or Linux — the legacy install.sh and bundled Compose files were deprecated in v0.130.0. SigNoz Cloud is the managed option, while RBAC, ingestion controls, custom retention, SSO, data residency, and support live in the Enterprise tier. Noz, the in-product AI teammate, is available only on SigNoz Cloud.
Categories:
Tags:
Frequently Asked Questions
What are the requirements for running self-hosted SigNoz?
You need Docker Engine 20.10 or newer with the Compose v2 plugin and at least 4 GB of memory allocated to Docker. Ports 8080 for the UI, 4317 and 4318 for OTLP ingestion, and optionally 8000 for the MCP server must be reachable. Linux and macOS are supported directly; on Windows you must run Docker Engine inside WSL 2 rather than Docker Desktop, because ClickHouse Keeper crashes there.
How do I install SigNoz, and can I still use the old install script?
Install the Foundry CLI with the signoz.io/foundry.sh script, write a casting.yaml describing the deployment, and run foundryctl cast. Foundry supports Docker Compose, Kubernetes, and Linux targets. The legacy install.sh script and the Docker Compose files bundled under deploy/ in the repository were deprecated in v0.130.0 and are no longer maintained, though a migration guide exists for moving an existing install to Foundry without losing data.
What monitoring signals and alert types does SigNoz cover?
It handles logs, metrics, traces, exceptions, and Kubernetes or host infrastructure telemetry. Alert rules can be built on metric thresholds, log patterns and counts, span conditions such as p99 latency or error counts, anomaly detection against historical baselines, and new or recurring exceptions. Notifications go to Slack, PagerDuty, Opsgenie, Microsoft Teams, email, or webhooks, with routing policies and planned maintenance windows to reduce fatigue, plus Terraform and Kubernetes operator support.
Does SigNoz work with AI coding agents?
Yes. The SigNoz MCP server lets AI clients query your telemetry, and Foundry can deploy it alongside the stack by enabling the mcp block in casting.yaml, which exposes it on port 8000. Clients such as Claude Code connect over HTTP with a SIGNOZ-API-KEY generated from a service account. Noz, the in-product AI teammate that investigates incidents and drafts dashboards, is a separate feature available only on SigNoz Cloud.
Which backends does SigNoz store telemetry in?
Traces, metrics, and logs are stored in ClickHouse, with ClickHouse Keeper providing coordination in single-node installs, and Postgres holds the metadata such as dashboards, alert rules, and user accounts. The query builder can also run PromQL or raw ClickHouse SQL, and dashboards accept any of those query modes on the same panel.
OpenAltFinder Score
Project Details
Repository details
Updated 11/10/2026, 9:01:11 AM
View RepositoryRepository activity
Get the latest open source alternatives delivered weekly.
Discover tools that give you freedom from proprietary software.
Compare SigNoz with
Similar open source alternatives
Beszel
Lightweight self-hosted server monitoring hub with historical data, Docker stats, and configurable alerts.
Datadog
Rootprint
Open-source, self-hosted log management with full-text search on object storage, OpenTelemetry ingestion, and team access control.
Datadog
ReJourney
Lightweight open-source observability tool with pixel-perfect session replay, real-time error detection, and journey mapping for web and mobile apps.
Sentry
Pingflare
Pingflare is a self-hosted uptime monitoring stack built natively for Cloudflare Workers. Get fast global checks, instant alerts, and a beautiful status page on your own domain.
Better Stack
OneUptime
Open-source observability platform combining uptime monitoring, incidents, on-call, status pages, logs, traces, and metrics.
Pingdom
UptimeKit
UptimeKit is a distributed monitoring and status page platform with fault-tolerant architecture. Watch services worldwide and keep users informed at a glance with clean public pages.
Better Stack