HyperDX – Observability Without the Datadog Price Tag

TL;DR

TL;DR: HyperDX is an open source observability platform that combines session replays, logs, traces, metrics, and error tracking in one affordable package — positioned as a Drop-in replacement for Datadog at a fraction of the cost.

Source and Accuracy Notes

• Product: hyperdx.io
• GitHub: hyperdxio/hyperdx
• Pricing: hyperdx.io/pricing
• Originally launched as DeploySentinel (YC S22) — HN Thread

What Is HyperDX?

HyperDX is an open source observability platform acquired by ClickHouse in 2024. It unifies session replays, logs, traces, metrics, and errors into a single dashboard — designed to help teams debug production issues faster without the enterprise price tag.

The core value proposition is cost reduction. Datadog and similar tools can cost thousands per month at scale. HyperDX positions itself as the open source alternative that doesn’t compromise on features.

The platform supports multiple programming languages: Node.js, Python, Java, Golang, Elixir, React Native, and more. It integrates with infrastructure providers like Kubernetes, AWS EC2, Vercel, Fly.io, Heroku, and Cloudflare.

Setup Workflow

Step 1: Create a Free Account or Self-Host

Cloud option (fastest):

# Sign up at https://app.hyperdx.io
# Get your ingestion API key from the dashboard

Self-hosted option:

# Clone the repository
git clone https://github.com/hyperdxio/hyperdx.git
cd hyperdx

# Follow the self-hosting guide in README
# Requires Docker, PostgreSQL, and ClickHouse

Step 2: Install the SDK

Node.js:

npm install @hyperdx/node-sdk

Python:

pip install hyperdx-python

Other languages: See the official documentation for your language of choice.

Step 3: Configure Instrumentation

Node.js example:

import { HdxSDK } from '@hyperdx/node-sdk';

const hdx = new HdxSDK({ apiKey: process.env.HYPERDX_API_KEY });

// Automatically captures errors, performance metrics, and logs

Step 4: Integrate with Existing Infrastructure

HyperDX provides integrations for:

• Kubernetes (via DaemonSet or Helm chart)
• AWS EC2 (via automatic detection)
• Vercel (via official integration)
• Fly.io (via automatic deployment detection)
• Cloudflare (via Workers integration)

# Kubernetes Helm values example
replicaCount: 1
image:
  repository: hyperdxio/hyperdx
  tag: latest
env:
  - name: HYPERDX_API_KEY
    value: "your-api-key"

Deeper Analysis

What Makes It Different

Most observability platforms started as single-purpose tools (APM, logging, or session replay) and expanded outward. HyperDX was built from the ground up as a unified platform, which means correlation between data types is a first-class feature rather than an afterthought.

The ClickHouse acquisition brings serious database muscle. ClickHouse is known for high-speed analytical queries on massive datasets, which translates to fast log searches and trace aggregation even at high volume.

Session Replay Integration

One standout feature is the ability to correlate a user’s session replay with server-side traces and logs. When an error occurs:

1. You see the session replay showing what the user experienced
2. The replay is automatically linked to the corresponding server-side traces
3. You can jump directly from a replay frame to the exact log line that caused the issue

This end-to-end correlation significantly reduces mean time to resolution compared to toggling between separate tools.

Pricing Model

HyperDX uses an open source core + cloud subscription model. The self-hosted version is free, while the cloud version offers:

• Generous free tier for small projects
• Usage-based pricing for larger scale
• SOC 2 Type II compliance for enterprise

Compare this to Datadog’s per-host or per-DPU pricing, which can escalate quickly for compute-heavy workloads.

Practical Evaluation Checklist

• Does it replace Datadog for most use cases? For teams primarily needing APM, logs, and session replays, yes. For specialized needs like network performance monitoring or security SIEM, you may still need point solutions.
• Is the self-hosted version production-ready? Yes — the open source version is actively maintained and used in production by various teams.
• How’s the performance at scale? ClickHouse backend means query performance stays fast even with billions of log entries.
• What’s the migration path from Datadog? HyperDX provides migration guides for common Datadog use cases. The SDKs are intentionally similar to reduce friction.

Security Notes

• SOC 2 Type II certification for the cloud product
• Data never leaves your infrastructure when self-hosted
• API keys are scoped to specific permissions
• Audit logs available on enterprise plans

FAQ

Q: Can I self-host HyperDX? A: Yes. The full platform is open source and can be self-hosted using Docker, PostgreSQL, and ClickHouse. The cloud version offers fully managed hosting if you prefer not to operate your own infrastructure.

Q: How does pricing compare to Datadog? A: HyperDX is significantly cheaper at scale. The open source version is free. Cloud pricing is usage-based with a generous free tier. Datadog can cost thousands monthly at enterprise scale.

Q: Does it support browser session replays? A: Yes. HyperDX has a browser SDK that captures session replays including console logs, network requests, and user interactions. These replays can be correlated with backend traces.

Q: What languages are supported? A: Node.js, Python, Java, Golang, Elixir, and React Native are officially supported. Community SDKs may exist for other languages.

Conclusion

HyperDX solves the observability cost problem that plagues growing engineering teams. By combining session replays, logs, traces, and metrics in one open source platform — backed by ClickHouse’s query performance — it offers a credible path off Datadog’s pricing without sacrificing functionality.

If you’re currently paying thousands monthly for observability and feeling the pain, HyperDX is worth evaluating. The self-hosted option lets you try it with zero cost, and the migration from Datadog is straightforward for common use cases.

Start at hyperdx.io or check the GitHub repository for the self-hosted deployment.