Query metrics - Automate.ax

axiom.queryMetrics runs MPL against an OpenTelemetry metrics dataset and returns Axiom’s metrics-v2 response. Discover metric names and tags first; the time window defaults to the previous 30 minutes.

automations/axiom-query-metrics.automation.ts

import { automation, onHttpRequest } from "automate.ax"
import { axiom } from "automate.ax/axiom"

export default automation("Query metrics", () => {
  onHttpRequest({ scope: "automation" })
  axiom.queryMetrics({
    mpl: "metrics:http.server.duration | align to 1m using avg",
    startTime: "now-15m",
    endTime: "now",
  })
})

Inputs

Input	Required	Default	Description
`mpl`	Yes	—	Non-empty MPL query.
`startTime`	No	`"now-30m"`	Query-window start as an Axiom time expression or timestamp.
`endTime`	No	`"now"`	Query-window end.
`edgeUrl`	No	`"https://us-east-1.aws.edge.axiom.co"`	Axiom query edge or an HTTPS `*.edge.axiom.co` origin.

Output

Returns Axiom’s metrics-v2 result as a JSON object. Its keys and series shape depend on the MPL query.

Permission and failures

The API token needs the dataset-level query|read capability for every metrics dataset referenced by the MPL query. Axiom rejects invalid MPL, inaccessible datasets, and invalid time windows. Request-rate limits return 429, while exhausted query-cost limits return 430. Automate.ax uses Axiom’s reset timing when available.