SigNoz

SigNoz pairs an OpenTelemetry collector with ClickHouse columnar storage to unify application performance monitoring, distributed traces, metrics, and logs in one self-hosted system. Instead of maintaining separate Prometheus, Jaeger, and Elasticsearch clusters, teams ingest standard OTLP telemetry into a single database. This architecture keeps production telemetry under your control while enabling fast analytical queries across high-cardinality traces and metrics without per-host fees.

Why SigNoz instead of Datadog APM

SigNoz centers observability on vendor-neutral standards rather than proprietary agent runtimes. In Datadog, teams install proprietary agents and libraries, binding metrics and traces to vendor-defined conventions and custom-metric pricing tiers. SigNoz relies entirely on OpenTelemetry. Services emit standard OTLP data into a shared ClickHouse schema, allowing on-call engineers to jump directly from a latency spike on an APM graph to span waterfalls, trace attributes, and correlated logs without tool switching or aggressive sampling.

The tradeoff is operational responsibility. Datadog delivers turnkey synthetic testing, continuous code profiling, automated anomaly detection, and managed infrastructure out of the box. SigNoz requires teams to deploy and operate ClickHouse, scale collector pipelines, and manage storage retention. For teams standardizing on Kubernetes and OpenTelemetry, trading managed convenience for predictable infrastructure costs and complete data residency is an intentional operational choice.

What it does well

• Native OpenTelemetry ingestion: Ingests standard OTLP traces, metrics, and logs directly, letting developers instrument applications once with open standards.
• ClickHouse storage: Handles high-cardinality attributes like user IDs and commit hashes with fast analytical queries and efficient data compression.
• Correlated diagnostics: Connects service latency charts, trace waterfalls, and log streams in a unified view to accelerate incident triage.
• Dual-query support: Enables DevOps engineers to build dashboards using PromQL syntax or query raw telemetry tables with ClickHouse SQL.
• Trace-derived exceptions: Captures runtime errors and stack traces from trace spans, tracking error rates without a separate monitoring agent.
• Trace funnels: Tracks multi-step request paths across distributed microservices to pinpoint drop-offs and downstream latency bottlenecks.