========================== A Docker image and example code for testing tagged metrics with Netuitive's StatsD backend.

1.1 Clone this repo and change into the netuitive-statsd-agent directory.

1.2 Build the Docker image:

$ sudo docker build --rm=true -t netuitive-statsd-agent .

1.3 Run a container from this image:

$ sudo docker run -d --name netuitive-statsd-agent -v /proc:/host_proc:ro -v /var/run/docker.sock:/var/run/docker.sock:ro -p 8125:8125/udp -d netuitive-statsd-agent

This will run the agent inside the container, exposing UDP port 8125. You'll need to determine your container's IP address in order to send metrics to it.

The config.js file will need to be modified to ensure a valid API key for Netuitive is in use:

{
    backends:["./backends/netuitive"],
    netuitive: {
        apiKey: "<API-KEY>",	<------ PUT YOUR API KEY HERE!
        apiHost: "api.app.netuitive.com",
        apiPort: 443,
        mappings: [
            {
      pattern: "(test.rob-egan)\\.(head.requests)\\.counter",
      element: {
        type: "SERVER",
        name: "$1",
        metric: {
          name: "$2"
        }
      }
    },
...

You can obtain the API Key by viewing the StatsD datasource you wish to use in the Netuitive UI

Two (overly simplified) code snippets have been provided that will create generic web services that will generate metrics and send them to the StatsD agent.

3.1 Python example

This example assumes you have Python 2.6 or greater installed, with the time, BaseHTTPServer, and statsd modules installed.

$ ./tests/webserver.py &

This will run a web server, listening on port 8000. You can hit the URL http://localhost:8000 in order to begin generating metrics.

3.2 Golang example

This example assumes you have Golang installed, and uses a thrid party Statsd client library [quipo/statsd]

$ go run ./tests/webserver.go &

This will run a webserver, listening on port 8000. Hit the URL http://localhost:8000/help for details on what paths are valid. Each valid path will generate metrics when hit.

3.3 Golang/Datadog Example

This example requires Golang plus the datadog-go statsd libraries. It was added to show a comparison between Datadog and Netuitive with regards to adding tags within the code being used to instrument metrics via StatsD.

$ go run ./tests/webserver_dd.go &

This example also assumes that a datadog-statsd agent is running.

This will create a web service on localhost, listening on TCP port 8000. The URL Path "/help" will explain details on what valid paths can be access, each of which produce various metrics.