HyperDX — Open Source Full-Stack Observability Platform

Introduction

HyperDX is an open-source observability platform that correlates logs, metrics, traces, errors, and session replays in a single unified view. Built on ClickHouse for fast queries and OpenTelemetry for vendor-neutral data collection, it helps developers resolve production issues faster by connecting frontend behavior to backend telemetry.

What HyperDX Does

• Correlates end-to-end traces with associated logs and metrics automatically
• Provides browser session replay linked to backend errors and traces
• Offers a search interface with full-text and structured querying across all signals
• Generates dashboards and alerts from any combination of telemetry data
• Supports OpenTelemetry-native ingestion for logs, traces, and metrics

Architecture Overview

HyperDX ingests telemetry via OpenTelemetry collectors and stores everything in ClickHouse for high-performance analytical queries. A Node.js API server handles correlation logic and serves the React frontend. MongoDB stores user and dashboard metadata. Session replay data flows through a separate capture SDK that links browser events to trace IDs.

Self-Hosting & Configuration

• Deploy with a single Docker Compose file (includes ClickHouse, MongoDB, API, and UI)
• Ingest data via standard OpenTelemetry SDKs and collectors
• Configure retention policies per signal type in the admin panel
• Supports SSO via OIDC for team access control
• Horizontal scaling possible by separating ClickHouse and the API tier

Key Features

• Unified search: query logs, traces, and metrics with one syntax
• Session replay with network waterfall and console logs linked to traces
• Automatic error grouping and pattern detection
• Custom dashboards with mixed signal types on a single chart
• Drop-in replacement for Datadog, Splunk, or New Relic with OTel compatibility

Comparison with Similar Tools

• Grafana + Loki + Tempo — modular stack requiring glue; HyperDX is integrated out of the box
• SigNoz — similar OTel-native approach; HyperDX adds session replay
• Datadog — commercial SaaS with usage-based pricing; HyperDX is self-hosted and free
• Highlight.io — also open source with replay; HyperDX uses ClickHouse for scale
• Elastic APM — heavier stack; HyperDX is lighter to operate

FAQ

Q: What OpenTelemetry SDKs are supported? A: Any language with an OTel SDK works. HyperDX provides optimized setup guides for Node.js, Python, Go, Java, and .NET.

Q: How much disk does ClickHouse consume? A: ClickHouse compresses aggressively; typical workloads see 10:1 compression ratios compared to raw log sizes.

Q: Can I use HyperDX alongside my existing Prometheus setup? A: Yes, HyperDX can ingest Prometheus remote-write metrics alongside OTel traces and logs.

Q: Is there a hosted/cloud version? A: HyperDX offers a managed cloud option, but the self-hosted edition is fully featured with no artificial limits.

Sources

• https://github.com/hyperdxio/hyperdx
• https://www.hyperdx.io/docs