Tail
Watch the platform live.
Stream logs as they land. Filter by service, level, or any structured attribute — the noise stays on the sidings.
Platform · Logs
Every line, on the record.
Structured logs, tied to the trace that wrote them. Tail live, search everything, and let ClickHouse keep the fare down.
Transfer
Every line knows its trace.
Logs carry trace context automatically. One click from the error line to the span that wrote it, the request around it, and the replay.
Search
Ask in attributes, not regexes.
Logs arrive over OTLP, structured. Query by service, level, user, or any field your app attaches — across everything you’ve kept, not just tonight’s buffer.
On this line
The parts you’d otherwise assemble.
OTLP native
Point your existing OpenTelemetry SDKs at Traceway. No vendor SDK, no collector required.
ClickHouse economics
A million events a day is roughly 2 GB a month at rest. Retention you can actually afford.
Alerts on patterns
New error pattern in the stream? It reaches Slack, Telegram, or a webhook before your users do.
Agents can read it
The agent skills query logs directly — “find the errors behind the checkout 500s” is a complete instruction.
All aboard
Put your logs on the line.
Point OTLP and tail within minutes — cloud or your own metal. MIT, full parity.