Obstack collects traces, metrics and logs without code changes, keeps them on servers you own, and shows them in Eagle and Grafana. One flat price a month, whatever the size of your fleet. No per-node licence, no per-gigabyte meter, no phone-home.
A Python application emits real metrics, logs and traces. Watch the signals in Grafana, then explore Eagle for the console, EagleSRE for incidents and EagleMCP for assistant access.
Every product speaks OpenTelemetry, so each one works with what you already run, and the next one slots in without a migration.
Distributed tracing and profiling, on its own.
Bee traces and profiles inside a microservice and between processes on the same host, then continues the same trace across hosts. HTTP, gRPC, TLS, MySQL, Redis and named Node.js awaits, no SDK. Exports OTLP to any OpenTelemetry collector.
The complete stack, built around Bee.
Bee and Beyla for collection, Vector for transport, GreptimeDB for storage, Grafana for dashboards, Prometheus Alertmanager for alert routing, and the Grafana MCP server for agents. Eagle provides the console and EagleMCP connects assistants. EagleSRE is in preview. Self-hosted, or run by us in the region you choose, at the same flat price. We configure it end to end; custom dashboards, alerts and migration are quoted and billed on time and materials. One support contract.
On-call schedules, escalation policies, incidents and an investigation agent, alongside Eagle. Respond from the console, Slack or Microsoft Teams, using the same identity.
Dashboards, Explore and alerting in Grafana, with the Grafana MCP server so agents query the same datasources your team does. In every plan, with unlimited users, not sold on its own.
How Grafana works hereConnect your MCP-compatible assistants to the Obstack Stack. Bring telemetry into the conversation alongside Eagle and EagleSRE.
Our own observability console, next to Grafana: PromQL, LogQL and TraceQL behind one search, Ask Eagle answering from your telemetry, and Serverless functions on an edge runtime.
What's in itEvery signal lands in the same store with the same labels, so a slow request, the host it ran on and the device that sent it can be read side by side.
Bee traces and profiles inside a microservice and on the host. Requests, database calls and hops between services become one distributed trace, with no SDK and no coordinator.
Hosts, containers and Kubernetes nodes report CPU, memory, disk and logs through Vector, while Bee maps which services talk to which.
Gateways and devices send OTLP into the same pipeline. Tell us about your fleet: email us.
eBPF agents on every node: Bee goes deep on the services that matter, Beyla goes broad on the rest.
Vector gathers OTLP, host metrics and logs on every node, and routes every signal onward.
Metrics, logs and traces in one database, queried with SQL and PromQL.
Dashboards and exploration on the Grafana your team already knows, plus the Grafana MCP server for agents. Eagle, our own console, sits beside it.
EagleSRE brings on-call and incident investigation together. EagleMCP connects assistants to the stack. EagleMCP is included; EagleSRE is in preview.
Telemetry goes to collectors and databases you run. Bee ships under a license key for the first month; after you sign the PolyForm Internal Use agreement, you get the source over git for as long as you subscribe. Air-gapped delivery is quoted as Enterprise, not on the list plans.
OTLP and Prometheus on the way in, SQL, PromQL and Grafana on the way out. If you ever leave, your data is already in formats every tool reads.
Everyone else licenses observability per node, so the invoice grows every time you add a machine. Obstack charges a flat monthly fee for the plan: twenty nodes and two thousand cost the same, and an incident that triples your traffic changes nothing.
Install the Max demo on your Kubernetes cluster with our Helm chart and a 30-day evaluation license. Use your own telemetry, then activate the same installation on self-hosted Go, Pro or Max. No card or automatic subscription.
