SigNoz

SigNoz is an open-source APM and observability platform for traces, metrics and logs, built on OpenTelemetry and run as a usage-billed cloud or self-hosted.

SigNoz shows engineers why a request was slow or failed by putting its trace, the logs it wrote and the surrounding metrics side by side. SigNoz Inc. builds it, and its founders named it after the gap between signal and noise in monitoring data. Every signal arrives through OpenTelemetry, so services instrumented once send data without a proprietary agent. The source code is public on GitHub. The APM view lists every service with its latency percentiles, error rate, request rate and Apdex score. The Service Map draws the services, databases and messaging systems that requests touch, built from traces with no manual setup. A trace opens as a flamegraph of spans, and one click moves to the logs emitted during it, or from a log line back to its trace. The Logs explorer pairs full-text search with field filters, saved views and live tail. Exceptions are grouped by service and type, while dashboards and alerts are written in the query builder, PromQL or ClickHouse SQL. SigNoz serves engineering, SRE and DevOps teams running microservices who want traces, logs and metrics in one tool instead of three. Teams already instrumented with OpenTelemetry point their collectors at it without new agents, and teams leaving Datadog have migration guides to follow. Buyers choose SigNoz Cloud, which SigNoz operates, or the open-source Community Edition, which they host themselves. Larger organisations sign an enterprise contract for a dedicated cloud, a deployment managed inside their own cloud account, or support for a self-hosted install. On-call engineers follow an alert on tail latency or error rate to the slow span and the log lines that explain it. Platform teams send host metrics, Kubernetes pod logs and AWS Lambda traces into the same store, so an incident reads as one timeline across services. Tail sampling in the OpenTelemetry Collector keeps the trace volume of the busiest services in check. Teams building on large language models add LLM observability, and the Noz AI teammate and an MCP server let AI agents query the same telemetry. SigNoz fits teams that want one place to answer why a request failed, on SigNoz Cloud or on their own servers.