← all SaaS
Self-host Honeycomb
Distributed tracing / observability ·
Category: observability & analytics
Honeycomb is the developer-loved tracing platform — high-cardinality event analysis, BubbleUp anomaly exploration, SLO management, all built around wide events. The self-hostable replacements all do trace ingest + query + alert; what they don't replicate is Honeycomb's signature query UX (BubbleUp, heatmaps over arbitrary fields).
Honeycomb pricing anchor: Free 20M events/mo; Pro $130/mo for 100M events; Enterprise custom.
- GitHub
- ★ 26.9k · last commit 1d ago · 1531 open issues
- License
-
MIT - Setup time
- 30min docker-compose (SigNoz + ClickHouse + collector)
- Monthly cost
- $30-60/mo VPS for production traffic.
Migration sketch. Use the official compose at signoz.io/docs/install/docker. Apps already shipping OTLP to Honeycomb just need a new endpoint — set `OTEL_EXPORTER_OTLP_ENDPOINT` to your SigNoz collector. SigNoz's services view, error tracking, and trace explorer cover the bulk of Honeycomb workflows; SLO support is built in. BubbleUp-style anomaly exploration is thinner — you'll write more queries by hand.
Good fit forTeams already using OpenTelemetry that want a single product covering traces + metrics + logs.
Weak atHoneycomb's BubbleUp / interactive explorer doesn't have a one-for-one match — query workflows are more SQL-shaped.
- GitHub
- ★ 22.8k · last commit today · 401 open issues
- License
-
Apache-2.0 - Setup time
- 10min docker run (`all-in-one`) for dev; 1h for production with Cassandra/Elasticsearch backend
- Monthly cost
- $5/mo VPS for dev; $50+/mo for production with Elasticsearch storage.
Migration sketch. `docker run --rm -d -p 16686:16686 -p 4317:4317 -p 4318:4318 jaegertracing/all-in-one`. Apps export OTLP to Jaeger directly (it's an OTel collector + storage + UI in one image). Migration from Honeycomb: change OTLP endpoint to Jaeger's. Trace search via service/operation/tag is the focus; metrics and logs are not in scope (Jaeger is traces-only).
Good fit forPure trace exploration — span search, dependency graphs, latency histograms — without the metric/log surface.
Weak atNo metrics or logs; no SLO management; you'll pair with Prometheus and Loki for the rest.
- GitHub
- ★ 5.25k · last commit 1d ago · 159 open issues
- License
-
AGPL-3.0 - Setup time
- 30min docker-compose (Tempo + Grafana + collector)
- Monthly cost
- $15-50/mo VPS — S3-backed for trace storage keeps cost predictable.
Migration sketch. Use the `docker-compose-tempo` template from grafana.com/docs/tempo. Apps export OTLP to Tempo via the collector; query traces in Grafana with TraceQL. Pair with Loki (logs), Mimir/Prometheus (metrics) for full observability. TraceQL is more powerful than Jaeger's tag-search but not as ergonomic as Honeycomb's query builder.
Good fit forTeams already invested in Grafana that want traces in the same pane as metrics and logs.
Weak atNo first-class anomaly-explore UX; you build dashboards manually.
In a terminal? npx os-alt honeycomb prints this table —
how the CLI works →