InfluxDB DWH Exporter

This application provides an HTTP endpoint that can deliver metrics from an InfluxDB in an aggregated form when called.

This can be used to provide an aggregated view of existing metrics or derived ones to an external tool (e.g. data warehouse or other databases).

Usage

The endpoint which is provided and through which the metrics can be retrieved/exported is /dwh

Examples

/dwh?token=my_secret - Returns the data of the last minute in minute resolution with an offset of one minute.

/dwh?token=my_secret&interval=15m - Returns the data of the last 15 minutes in a 15-minute resolution with an offset of one minute.

/dwh?token=my_secret&interval=15m&range=60m&offset=5m - Returns the data of the last 60 minutes in a 15-minute resolution with an offset of five minutes.

/dwh?token=my_secret&interval=1m&start=1587023340000&end1587023400000 - Returns the data between defined the timestamps in a 1-minute resolution.

Configuration

The configuration of the provided metrics can be done via an application.yml file, which can be put in the working directory of the application.

The following code is an example for the application.yml. For a full list of available configurations, see the default application.yml:

server:
  # the port of the server
  port: 8080

spring:
  influx:
    # URL of the InfluxDB
    url: http://localhost:8086
    # username for the InfluxDB
    user: ""
    # password for the InfluxDB
    password: ""

dwh:
  # secret to access the /dwh endpoint
  token: my_secret

  # the metrics and related queries to use and export
  metrics:
    - name: my-metric|${service}|${http_path}
      query: |
        SELECT SUM("sum") / SUM("count")
        FROM "inspectit"."autogen"."http_in_responsetime"
        GROUP BY time(${interval}), "service", "http_path" FILL(null)

Metrics Configuration

A list of metrics that will be provided via the HTTP endpoint can be specified in dwh.metrics. Two parameters must be defined for each metric: name and query.

name is the name of the resulting "metric view". The name can be parameterized using the tags of the underlying metric, as shown in the example above using ${service} and ${http_path}.

query is the InfluxDB query that is executed to retrieve the corresponding metric from InfluxDB. Here, the variable ${timeFilter} should be used in the WHERE clause. Also, {interval} should be used as "GROUP BY time" interval!

It is important that exactly the tags used for parameterization in name are also used in the GROUP BY-clause!

When the endpoint provided by the application is called, the exporter executes all configured queries and returns the corresponding result.

Health Endpoint

The application provides an endpoint under `/actuator/health that can be used to check if the application is running and the InfluxDB is available.

By default, only a simple status is provided:

{
    "status":"UP",
}

However, with the following configuration, additional information, such as status of the InfluxDB connection, can be obtained via the health endpoint:

management:
   endpoint:
      health:
        show-details: "ALWAYS"

Using this configuration, the health endpoint provides the following result:

{
    "status":"UP",
    "components": {
        "influxDb": {
            "status":"UP"
        }
    }
}