Telemetry - script API

StackState Self-hosted v4.6.x

This page describes StackState version 4.6.

Go to the documentation for the latest StackState release.

Function query

A telemetry query is a conjunction of equality conditions. For example, name = 'system.load.norm.15' and host='localhost'. There are several builder methods available that help to refine the query time range, limit the number of points returned, or set a metric field.

Telemetry.query(dataSourceName: String, query: String)

Telemetry queries only support metric queries. If you need event queries, please enter a feature request at support.stackstate.com

Args:

• dataSourceName - name of the data source.

• query - set of equality conditions.

Returns:

AsyncScriptResult[TelemetryScriptApiQueryResponse]

Builder methods:

• .groupBy(fieldName: String) - optional. Used to return grouped results from Elasticsearch. Requires .aggregation() to be used. If there is no aggregation, a plain metric stream will be returned.

• aggregation(method: String, bucketSize: String) - returns aggregated telemetry using method and bucketSize. See the available aggregation methods.

• start(time: Instant) - sets the start time of the query, for example -3h.

• end(time: Instant) - sets the end time of the query, for example -1h.

• window(start: Instant, end: Instant) - sets query time range. Use only start to get all telemetry up to now or only end to get all telemetry up to an instant in time.

• limit(points: Int) - limits the number of points returned, applicable to none aggregated queries.

• metricField(fieldName: String) - optional, but may be required for some data sources. Sets a field that holds metric value.

• compileQuery() - returns the telemetry query that was created with this function and the builder methods. After this builder method no more builder methods can be called.

Examples:

• Get metrics aggregated using Mean with bucket size 15 minutes and grouped by the field host:

Query

Telemetry
  .query("StackState Multi Metrics", "")
  .groupBy("host")
  .metricField("jvm_threads_current")
  .start("-15m")
  .aggregation("mean", "15m")

Example JSON output

{
  "_type": "TelemetryMetricQueryResponse",
  "query": {
    "_type": "TelemetryMetricQuery",
    "aggregation": {
      "bucketSize": 900000,
      "method": "MEAN"
    },
    "conditions": [
      {
        "key": "name",
        "value": "jvm_threads_current"
      }
    ],
    "dataSourceId": 277422153298283,
    "endTime": 1643294849765,
    "groupBy": {
      "fields": ["host"]
    },
    "includeAnnotations": false,
    "metricField": "stackstate.jvm_threads_current",
    "startTime": 1643293949765
  },
  "response": {
    "_type": "MetricData",
    "result": [
      {
        "_type": "TimeSeries",
        "annotations": [],
        "id": {
          "_type": "TimeSeriesId",
          "groups": {
            "host": "sts-kafka-to-es-multi-metrics_generic-events_topology-events_state-events_sts-events"
          }
        },
        "points": [[1643293800000, 49.46666666666667]],
        "tags": []
      },
      {
        "_type": "TimeSeries",
        "annotations": [],
        "id": {
          "_type": "TimeSeriesId",
          "groups": {
            "host": "sts-kafka-to-es-trace-events"
          }
        },
        "points": [[1643293800000, 44.46666666666667]],
        "tags": []
      },
      {
        "_type": "TimeSeries",
        "annotations": [],
        "id": {
          "_type": "TimeSeriesId",
          "groups": {
            "host": "sts-receiver"
          }
        },
        "points": [[1643293800000, 50.266666666666666]],
        "tags": []
      },
      {
        "_type": "TimeSeries",
        "annotations": [],
        "id": {
          "_type": "TimeSeriesId",
          "groups": {
            "host": "sts-server"
          }
        },
        "points": [[1643293800000, 179.53333333333333]],
        "tags": []
      }
    ]
  }
}

• Get raw metric by query

  Telemetry
    .query("StackState Metrics", "name='system.load.norm' and host='host1'")
    .metricField("value")

• Get metric aggregated using Mean with bucket size 1 minute:

  Telemetry
    .query("StackState Metrics", "name='system.load.norm' and host='host1'")
    .metricField("value")
    .aggregation("99th percentile", "1m") // get 99th percentile of each minute

• Query metrics starting 3 hours ago till now:

  Telemetry
    .query("StackState Metrics", "name='system.load.norm' and host='host1'")
    .metricField("value")
    .start("-3h") // starting from 3 hours ago

• Query metrics starting beginning of the data till last hour ago:

  Telemetry
    .query("StackState Metrics", "name='system.load.norm' and host='host1'")
    .metricField("value")
    .end("-1h") // ending 1 hour ago

• Query metrics within time range starting 3 hours ago up to 1 hour ago:

  Telemetry
    .query("StackState Metrics", "name='system.load.norm' and host='host1'")
    .metricField("value")
    .window("-3h", "-1h") // from 3 hours ago to 1 hour ago

• Query metrics from field "value" and limits points returned:

  Telemetry
    .query("StackState Metrics", "name='system.load.norm' and host='host1'")
    .metricField("value")
    .limit(100)