Application metrics via OTel
Emit Counter, Gauge, and Histogram metrics through the OpenTelemetry SDK you already have; tags become facet filters.
Platform · Metrics
Measure what matters, without the bill shock.
Application metrics via OpenTelemetry, automatic server metrics, and flexible widget dashboards — all included, with no per-metric billing and no surprise overages.
Service facts
What the meter says — and what it doesn't
The posted rates for metrics on Traceway.
- 60 shost scrape interval
- The Traceway OTel Agent collects CPU, memory, load, disk, filesystem, and network via the upstream hostmetricsreceiver.
- 0events charged per metric
- Custom metrics never count toward your event limit — only issues, HTTP requests, and background tasks do.
- 3types ingested as-is
- Counter, Gauge, and Histogram over OTLP/HTTP and OTLP/gRPC, with units and dimensional tags preserved.
No surge pricing
The meter doesn't run on metrics.
Emit thousands of custom metrics without watching a bill. The same generosity applies in your own yard: MIT licensed, full feature parity with Cloud, no asterisks.
/api/otel/v1/metricsPoint any OpenTelemetry SDK at it — OTLP/HTTP and OTLP/gRPC both supported natively.
- App metrics
- OTel-nativeDimensional tags become facet filters in the dashboard — chart signups by plan, region, or tenant.
- Host agent
- One-line installLinux (systemd), macOS (launchd), or Windows; set TRACEWAY_TOKEN and the installer wires the service.
- Per-process metrics
- Opt-inTRACEWAY_PROCESS_NAMES when you want them; Go runtime metrics come from the Go client SDK.
- Log tailing
- Also opt-inThe same agent tails any log files you point it at.
- Templates
- 4 starter packsOTel host agent, Go runtime, ClickHouse, and DuckDB dashboards install from the marketplace.
On this line
Two kinds of metrics, one bill of none.
Application metrics over OTLP, host metrics from a one-curl agent, and dashboards that travel across projects.
Host metrics in one curl
The Traceway OTel Agent is a pre-built collector distribution that scrapes host metrics every 60 seconds and ships them over OTLP/HTTP — no config file, no code changes.
Dashboards as documents
Each dashboard is a plain JSON document: edit it as code, keep it in your repo, apply it across every project, or import the Grafana dashboards you already have.
No per-metric billing
Metrics are included at no additional event cost, so thousands of custom metrics are a non-event on the invoice.
Next stop
Ship metrics in 5 minutes.
Application and server metrics, included on every plan. No per-metric billing.