Hyperdx

HyperDX unifies logs, traces, metrics, and session replay in one platform — instrument with OpenTelemetry, ship telemetry to HyperDX, and search across all signal types simultaneously. Clicking a log line jumps to the correlated trace; clicking a trace shows the correlated logs. For agents: instrument LLM applications with OTel, get correlated visibility across agent steps (trace) and system logs (log), and set alerts on error rate or latency thresholds. ClickHouse backend handles high-throughput telemetry at low cost. Self-hosted (Docker) or HyperDX Cloud. Confidence is docs-derived.

OTel instrumentation: OTEL_EXPORTER_OTLP_ENDPOINT="http://hyperdx-host:4318" + OTEL_SERVICE_NAME="my-service" — standard OTel setup sends traces, metrics, and logs. Browser instrumentation: @hyperdx/browser SDK for frontend session replay. Node.js auto-instrument: npx @hyperdx/node-opentelemetry — auto-instruments Express/Fastify/Next.js. Python: standard OTel SDK with OTLP exporter. HyperDX API: GET /api/v1/logs queries logs. GET /api/v1/traces queries traces. REST API key for programmatic access. Alert rules: POST /api/v1/alerts creates threshold alert. Webhook: configure Slack/PagerDuty webhook for alert delivery. Search syntax: uses Lucene-style query syntax — service:my-service AND level:error.

Self-hosted: no external auth required — deploy behind your own auth proxy. HyperDX Cloud: email/Google SSO; API keys for REST API. API key: Authorization: Bearer {api_key}. HTTPS enforced for Cloud; configure at reverse proxy for self-hosted. OTel endpoint: optional auth header (OTEL_EXPORTER_OTLP_HEADERS="authorization=Bearer ..."). RBAC: team roles in HyperDX Cloud. Data: self-hosted keeps all telemetry on your infrastructure.

ClickHouse: high-throughput ingestion (millions of events/second on modest hardware). Retention: configurable per signal type. Self-hosted resource: ClickHouse needs ≥4GB RAM for production. Docker Compose: suitable for dev/small teams; Kubernetes for production scale. Correlated search: log-trace correlation via trace_id in log attributes — requires consistent trace context propagation. Alert latency: ~1 minute evaluation window. HyperDX Cloud uptime: status.hyperdx.io.

www.hyperdx.io/docs covers OTel instrumentation, self-hosting guide, Docker Compose setup, search query syntax, alert configuration, and browser SDK. Getting started: docker compose up — HyperDX running in 5 minutes; instrument first app in 10 minutes. HyperDX Cloud: free tier (5GB/month). Self-hosted: fully free, ClickHouse storage. GitHub: hyperdxio/hyperdx (7k+ stars). Community via Discord (active). Good documentation; unified log-trace correlation guide is well-illustrated.