Axiom.Dataset reference

Annotation

Source: src/Axiom/Annotation.ts

An Axiom annotation — a vertical marker overlaid on charts to flag a deploy, incident, feature flag flip, or any other point/range event you want correlated with telemetry.

Annotations are scoped to one or more datasets. Use a single time for a point marker or time + endTime for a range. The type (e.g. "deploy", "incident") groups markers visually in the UI.

Although typically created at deploy/release time (out-of-band of regular IaC), modelling them as resources makes per-environment annotation history reproducible.

Annotation: Creating an Annotation

Point-in-time deploy marker

yield* Axiom.Annotation("deploy-1.2.3", {
  type: "deploy",
  title: "Release 1.2.3",
  description: "https://github.com/acme/app/releases/tag/v1.2.3",
  datasets: ["my-app-traces", "my-app-logs"],
  time: new Date().toISOString(),
  url: "https://github.com/acme/app/releases/tag/v1.2.3",
});

Incident time-range

yield* Axiom.Annotation("inc-2026-04-27", {
  type: "incident",
  title: "Database failover",
  datasets: ["my-app-traces"],
  time:    "2026-04-27T18:05:00Z",
  endTime: "2026-04-27T18:32:00Z",
  url: "https://incident.io/incidents/abc123",
});

Dataset

Source: src/Axiom/Dataset.ts

An Axiom dataset — the top-level container that stores events, logs, traces, or metrics. Pick a kind up-front: it determines schema and how the data is shown in the UI, and cannot be changed after creation (changing it triggers a replacement, which deletes the data).

Datasets expose their edge deployment and Axiom’s OTLP/HTTP endpoints (otelTracesEndpoint, otelLogsEndpoint, otelMetricsEndpoint) as output attributes so you can inject them into a Worker / Lambda’s env vars for OpenTelemetry shipping. The bearer token is not stored in resource state — supply Authorization: Bearer <AXIOM_TOKEN> separately at runtime.

Dataset: Creating a Dataset

Logs dataset with 30-day retention

const logs = yield* Axiom.Dataset("app-logs", {
  name: "my-app-logs",
  kind: "otel:logs:v1",
  description: "Application logs from prod workers",
  retentionDays: 30,
  useRetentionPeriod: true,
});

Separate datasets per OTEL signal

const traces  = yield* Axiom.Dataset("traces",  { name: "app-traces",  kind: "otel:traces:v1"  });
const logs    = yield* Axiom.Dataset("logs",    { name: "app-logs",    kind: "otel:logs:v1"    });
const metrics = yield* Axiom.Dataset("metrics", { name: "app-metrics", kind: "otel:metrics:v1" });

Dataset: Shipping OTEL data

yield* Cloudflare.Worker("api", {
  vars: {
    OTEL_EXPORTER_OTLP_TRACES_ENDPOINT: traces.otelTracesEndpoint,
    OTEL_EXPORTER_OTLP_LOGS_ENDPOINT:   logs.otelLogsEndpoint,
    // Bearer token must come from a secret store, not the dataset state.
    OTEL_EXPORTER_OTLP_HEADERS:
      `Authorization=Bearer ${env.AXIOM_TOKEN},X-Axiom-Dataset=${traces.name}`,
  },
});

VirtualField

Source: src/Axiom/VirtualField.ts

An Axiom virtual field — a saved APL expression that appears as a derived column on a dataset at query time. Use these to standardise common computations (status classes, latency buckets, parsed JSON paths) so dashboards and monitors don’t have to redefine them.

Bound to a single dataset; changing the dataset triggers a replacement.

VirtualField: Creating a Virtual Field

HTTP status class (e.g. 200 → “2xx”)

yield* Axiom.VirtualField("status-class", {
  dataset: "my-app-traces",
  name: "status_class",
  description: "HTTP response class bucket",
  expression: 'strcat(tostring(toint(status / 100)), "xx")',
  type: "string",
});

Latency bucket in seconds

yield* Axiom.VirtualField("latency-bucket", {
  dataset: "my-app-traces",
  name: "latency_bucket_s",
  expression: "bin(duration_ms / 1000.0, 0.5)",
  type: "number",
  unit: "s",
});