{"slug": "show-hn-pitot-opentelemetry-in-one-binary-with-parquet-duckdb-and-mcp", "title": "Show HN: Pitot – OpenTelemetry in One Binary, with Parquet, DuckDB, and MCP", "summary": "Labstack released Pitot, a pre-release open-source observability tool that ingests OpenTelemetry traces, logs and metrics into a single binary, stores them locally as Parquet files and queries them with DuckDB, with no external storage services required. Pitot exposes an MCP server so external AI agents can use its typed query tools, alongside a built-in investigation agent that answers plain-language questions about telemetry when configured with an Anthropic or OpenAI API key. Labstack said the browser client, built-in agent, MCP tools, dashboards and alerts are preview capabilities and that the pre-release software changes without migration paths.", "body_md": "Pitot (pronounced **PEE-toh**).\n\n**Observability your agents can operate.** Send OpenTelemetry to one binary, then investigate with\nPitot's built-in agent or connect your own over MCP.\n\n*Current Pitot interface with deterministic offline sample telemetry.*\n\nPitot keeps OpenTelemetry traces, logs, and metrics in one binary and one data directory, with no\nexternal storage services to operate. Pitot is pre-release and changes without migration paths; the browser\nclient, built-in agent, MCP tools, dashboards, and alerts are preview capabilities\n([status](https://pitot.dev/docs/reference/capabilities)). Start with the\n[documentation](https://pitot.dev) or run it below.\n\n- OTLP ingest over gRPC and HTTP, with local Parquet storage and DuckDB analytics.\n- Explore and PitotQL for querying telemetry without an AI provider.\n- An investigation agent that answers questions about your telemetry in plain language.\n- An MCP server for external agents using the same typed query tools.\n- Dashboards across traces, logs, and metrics, and alerts on service error rate, latency, throughput, and log volume.\n\n```\ndocker run --name pitot --restart unless-stopped -p 7520:7520 \\\n  -v pitot-data:/var/lib/pitot/data \\\n  -e PITOT_AUTH_CODE_SECRET=$(openssl rand -hex 32) \\\n  labstack/pitot:latest\n```\n\nOpen the one-time setup URL printed by the container and create the first administrator. Pitot\ndisplays the ingest token exactly once, unless `PITOT_INGEST_TOKEN` already set it; save it with your collector secrets. For later sign-in\nwithout SMTP, mint a 15-minute, single-use link against the running container's control database:\n\n```\ndocker exec pitot pitot login-link admin@example.com\n```\n\nThe UI, API, MCP, and both OTLP transports share port `7520`. See\n[installation](https://pitot.dev/docs/get-started/install) for image defaults and bind-mount ownership, and\n[first boot](https://pitot.dev/docs/get-started/first-boot) for setup and SMTP login.\n\nInstall a checksum-verified release for Linux or macOS, amd64 or arm64:\n\n```\ncurl -fsSL https://raw.githubusercontent.com/labstack/pitot/main/scripts/install.sh | sh\nexport PITOT_AUTH_CODE_SECRET=$(openssl rand -hex 32)\npitot\n```\n\nSet `PITOT_VERSION` to a [release tag](https://github.com/labstack/pitot/releases), such as `v2026.11.0-alpha`, to pin a release, and `PITOT_PREFIX` to choose the installation\ndirectory. Native installs bind `127.0.0.1:7520`; set `PITOT_ADDR=:7520` for remote access, then\nfollow the same first-boot setup.\n\nA standard OTLP/HTTP exporter can use the saved ingest token:\n\n```\nexport OTEL_EXPORTER_OTLP_PROTOCOL=http/protobuf\nexport OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:7520\nexport OTEL_EXPORTER_OTLP_HEADERS=\"authorization=Bearer%20$INGEST_TOKEN\"\n```\n\nFor a Collector on the same private container network, the forwarding side is:\n\n```\nexporters:\n  otlp_http/pitot:\n    endpoint: http://pitot:7520\n    headers:\n      Authorization: \"Bearer ${env:INGEST_TOKEN}\"\n\nservice:\n  pipelines:\n    traces:  { receivers: [otlp], exporters: [otlp_http/pitot] }\n    metrics: { receivers: [otlp], exporters: [otlp_http/pitot] }\n    logs:    { receivers: [otlp], exporters: [otlp_http/pitot] }\n```\n\nUse an existing `otlp` receiver, set `INGEST_TOKEN`, and replace `pitot` with the Collector's\nreachable private hostname. OTLP/gRPC uses `localhost:7520` locally; see [send\ntelemetry](https://pitot.dev/docs/get-started/send-telemetry) for TLS and HTTP/2 setup.\n\nSelect `PITOT_AI_PROVIDER=anthropic` and inject `ANTHROPIC_API_KEY`, or select\n`PITOT_AI_PROVIDER=openai` and inject `OPENAI_API_KEY`, then restart Pitot; the selected provider\nrequires its key at startup. With no provider, chat is disabled and the other features still work.\nSee [AI settings](https://pitot.dev/docs/reference/configuration/ai).\n\nSet `PITOT_PUBLIC_URL` to your stable, externally reachable HTTPS origin, then point a remote MCP\nclient at that origin's `/mcp` endpoint. The client discovers and registers through OAuth; you\nauthorize its scopes in the browser. Pitot serves MCP `2026-07-28` and, through the `initialize`\nhandshake, `2025-11-25` and `2025-06-18` over Streamable HTTP. See [connect over\nMCP](https://pitot.dev/docs/guides/mcp) for setup and permissions.\n\n| If you use | Where Pitot differs | \n|---|---|\n| **Grafana with Loki, Tempo, and Mimir** | That stack keeps a service and a query language per signal, plus object storage underneath. Pitot keeps one process, one data directory, and one typed contract across all three signals, at the cost of the horizontal scale those components are built for. | \n| **SigNoz** | Both are OTLP-native and self-hosted. SigNoz composes a collector, ClickHouse, and query services; Pitot compiles ingest, authoritative Parquet storage, indexed trace reads, DuckDB analytics, alerting, and the browser client into one binary. | \n| **Jaeger** | Jaeger covers traces and expects a storage backend you run separately. Pitot ingests traces, logs, and metrics into the same store, with nothing else to deploy. | \n| **Prometheus with Grafana** | Prometheus pulls metrics and is excellent at them. Pitot accepts pushed OTLP for all three signals and is built around investigating a specific incident rather than maintaining long-range metric series. | \n| **Datadog** ,**Honeycomb** ,**Grafana Cloud** | Those are managed services: someone else runs the storage, the scaling, and the upgrades, and your telemetry leaves your network to get there. Pitot is a binary you run, on data that stays on your disk. | \n| **An OpenTelemetry Collector piped into ClickHouse** | The same shape, assembled by hand: collector, database, dashboards, and the glue between them. Pitot is that assembly as one program, with an agent and an MCP server already wired to the same query contract. | \n\nPitot is a single node. It has no clustering, no replication, and no object tier; a deployment that outgrows one machine's disk and CPU has outgrown Pitot.\n\nOne process contains ingest, storage, query, alerts, the agent, MCP, and the React client. Browser and agent clients share a typed observability contract.\n\nParquet holds telemetry, DuckDB query state is rebuildable, and SQLite holds transactional product\nstate. See [how storage works](https://pitot.dev/docs/concepts/storage) for durable ingest,\nindexed reads, and compaction while reads continue.\n\nConfiguration is environment-only: built-in defaults followed by nonempty process environment values, validated before data files or listeners open. Settings are read once at startup; every change requires a restart.\n\n| Environment variable | Default | Purpose | \n|---|---|---|\n| `PITOT_ADDR` | `127.0.0.1:7520` | UI, API, MCP, and both OTLP transports | \n| `PITOT_DATA_DIR` | `./data` | Parquet, query state, and control SQLite | \n| `PITOT_AUTH_MODE` | `local` | Login link or SMTP, or `oidc` | \n| `PITOT_AUTH_CODE_SECRET` | — | Required in local mode and for OIDC break-glass links, 32+ characters | \n| `PITOT_AI_PROVIDER` | — | Empty disables chat; `anthropic` or`openai` enables it | \n| `ANTHROPIC_API_KEY` | — | Required for Anthropic chat | \n| `OPENAI_API_KEY` | — | Required for OpenAI chat | \n| `PITOT_RETENTION_DAYS` | `30` | Telemetry retention window | \n| `PITOT_MCP_ENABLED` | `true` | Serve the MCP endpoint at `/mcp` | \n\nEvery secret accepts `<NAME>_FILE` instead of its value; see the [settings\nreference](https://pitot.dev/docs/reference/configuration) for validation and [runtime secret\ndelivery](https://github.com/labstack/pitot/blob/main/docs/operations.md#runtime-secret-delivery) for deployment examples.\n\nPitot sizes DuckDB automatically; measured workloads can override memory, connections, and threads\nas described in [DuckDB sizing](https://pitot.dev/docs/reference/configuration#duckdb-sizing).\n\nFor TLS, health checks, backups, upgrades, retention, and recovery, see the [operator\nrunbook](https://github.com/labstack/pitot/blob/main/docs/operations.md).\n\nPitot v2026.11.0-alpha accepted **507,311 mixed OpenTelemetry items/s**\nover 306.8 seconds on a Hetzner Cloud ccx33 (8 dedicated vCPUs, 32 GB RAM), with\ntelemetrygen on a separate ccx33. The ingest run on 6 November 2026 passed every\nquality gate. See the [performance methodology](https://pitot.dev/docs/concepts/performance).\n[Pitot Bench](https://github.com/labstack/pitot-bench) reproduces the measurement on your hardware.\n\n- **Go 1.27.3 and a C/C++ toolchain** on native Linux or macOS, amd64 or arm64.\nBuild with`just build` or`bash scripts/with-duckdb.sh go build ./cmd/pitot` ;\nthe wrapper supplies DuckDB's pinned headers, static libraries, and CGO flags.\nThe first build needs network access to download the checksum-verified engine.\n- **[Bun](https://bun.sh)** — compiles the browser assets\n- **Node.js 24.19.0** — runs development helpers and builds the documentation site\n- **[just](https://just.systems)** — task runner\n- A 32-character **authentication code secret**\n\n```\ngit clone https://github.com/labstack/pitot.git\ncd pitot\njust install   # browser dependencies and git hooks\njust build     # browser assets, then the binary\n```\n\nRun it with the minimum configuration:\n\n```\nexport PITOT_AUTH_CODE_SECRET=$(openssl rand -hex 32) # must be 32+ characters\n\n./bin/pitot\n```\n\nOpen [http://localhost:7520](http://localhost:7520) and follow first-boot setup. SMTP and chat are optional; without SMTP,\nrun `./bin/pitot login-link admin@example.com` with the same configuration and data directory.\n\nRun `just` to list recipes, `just check` for the full gate, `just test` for Go tests, and `just ui`\nto rebuild embedded browser assets. See [CONTRIBUTING.md](https://github.com/labstack/pitot/blob/main/CONTRIBUTING.md) for environment setup,\ngenerated assets, hooks, and project layout.\n\nVersions are CalVer — `vYYYY.M.N[-alpha|-beta|-rc]`, numbered from 0 within each month across all\nmaturity suffixes. Pitot is alpha (for example, `v2026.11.0-alpha`); GA has no suffix.\nContainer tags omit the leading `v`, and every release moves `latest`. Tagged releases publish\nidentical manifests to Docker Hub (`labstack/pitot`) and GHCR (` ghcr.io/labstack/pitot`); GHCR is\ncanonical and also carries development images.\n\n| Image tag | Points at | \n|---|---|\n| `labstack/pitot:latest` | the newest release, mirrored on Docker Hub | \n| `labstack/pitot:2026.11.0-alpha` | that exact release, mirrored on Docker Hub | \n| `ghcr.io/labstack/pitot:latest` | the newest release | \n| `ghcr.io/labstack/pitot:2026.11.0-alpha` | that exact release | \n| `ghcr.io/labstack/pitot:main` | the tip of `main` | \n| `ghcr.io/labstack/pitot:sha-<commit>` | one specific commit | \n\nSee [docs/release.md](https://github.com/labstack/pitot/blob/main/docs/release.md) for platforms, checksums, provenance, and verification.\n\nIssues and pull requests are welcome; follow [CONTRIBUTING.md](https://github.com/labstack/pitot/blob/main/CONTRIBUTING.md) and run `just check`\nand `just test-race` before opening a pull request. Report security issues through\n[SECURITY.md](https://github.com/labstack/pitot/blob/main/SECURITY.md).\n\nThis repository builds Pitot without another repository; LabStack's deployment configuration,\nuptime monitoring, and public demo instance are not included. The documentation site in\n[`site/`](https://github.com/labstack/pitot/blob/main/site) is included, with settings pages generated from the same `internal/config` type the\nloader binds and checked by `just check`.\n\n[Apache-2.0](https://github.com/labstack/pitot/blob/main/LICENSE) © LabStack LLC. See [NOTICE](https://github.com/labstack/pitot/blob/main/NOTICE) and\n[THIRD_PARTY_NOTICES](https://github.com/labstack/pitot/blob/main/THIRD_PARTY_NOTICES) for attribution, and\n[TRADEMARK](https://github.com/labstack/pitot/blob/main/TRADEMARK.md) for use of the Pitot name and logo.", "url": "https://wpnews.pro/news/show-hn-pitot-opentelemetry-in-one-binary-with-parquet-duckdb-and-mcp", "canonical_source": "https://github.com/labstack/pitot", "published_at": "2026-11-18 23:55:35+00:00", "updated_at": "2026-11-19 00:16:12.294530+00:00", "lang": "en", "topics": ["ai-agents", "agent-protocols", "ai-tools", "developer-tools", "mlops"], "entities": ["Labstack", "Pitot", "OpenTelemetry", "DuckDB", "Parquet", "Anthropic", "OpenAI", "MCP"], "also_reported_by": [], "takeaway": "Pitot ingests OTLP over gRPC and HTTP into local Parquet storage queried by DuckDB, and exposes the same typed query tools to external agents through an MCP server at the /mcp endpoint with OAuth scope authorization.", "recommendation": null, "alternates": {"html": "https://wpnews.pro/news/show-hn-pitot-opentelemetry-in-one-binary-with-parquet-duckdb-and-mcp", "markdown": "https://wpnews.pro/news/show-hn-pitot-opentelemetry-in-one-binary-with-parquet-duckdb-and-mcp.md", "text": "https://wpnews.pro/news/show-hn-pitot-opentelemetry-in-one-binary-with-parquet-duckdb-and-mcp.txt", "jsonld": "https://wpnews.pro/news/show-hn-pitot-opentelemetry-in-one-binary-with-parquet-duckdb-and-mcp.jsonld"}}