Setting up the monitoring with Datadog

This section explains, in case of a Kubernetes deployment, how to set up the collection by Datadog of the metrics provided through the SonarQube’s Web API (Openmetrics format). In the following, we assume that you are installing both SonarQube and Datadog to a Kubernetes cluster via the corresponding Helm charts.

Introduction

To set up Datadog to monitor SonarQube, you have to specify an annotation in SonarQube’s Helm chart related to Datadog. Since Datadog doesn’t understand the native Prometheus’ Bearer authentication annotation, you cannot use it. Instead, you can specify an annotation that will manage a Datadog configuration file. 

The illustration below shows the setup and monitoring process:

• When you install the SonarQube’s Helm chart in the Kubernetes cluster, the chart deploys the Datadog configuration file on the SonarQube Pod.
• The Datadog agent then:
  1. Reads the Datadog configuration file.
  2. Authenticates to the SonarQube’s Web API endpoint and pulls the Prometheus metrics from the endpoint.
  3. Pushes the metrics to the Datadog dashboard.

Setting up the Datadog authentication to the Web API endpoint

You need to create a secret containing the monitoring passcode and then mount that secret in the Datadog agent. To do so, add the code below to the values.yaml file of the Datadog’s Helm chart. In this example, we mount the subkey passcode from the datadog-api-secret secret into /etc/secret-volume.

agents:

 volumes:

 - name: secret-volume

   secret:

     secretName: datadog-api-secret

     items:

     - key: passcode

       path: passcode

 volumeMounts:

 - name: secret-volume

   mountPath: /etc/secret-volume

Specifying the annotation for the Datadog agent

Add the code below to the values.yaml file of the SonarQube’s Helm chart. Note that: 

• This example corresponds to the example shown in Setting up the authentication to the Web API endpoint: you must adapt the reader and writer sections to your values.
• If a webcontext is used in the path at which to serve SonarQube then you must add it to the openmetrics_endpoint. For example, if the /sonarqube web context were used here then we would have:
  "openmetrics_endpoint": "http://%%host%%:9000/sonarqube/api/monitoring/metrics"

# Set annotations for pods

annotations:

 #ad.datadoghq.com/<EXPECTED_POD_NAME>.checks  

 ad.datadoghq.com/sonarqube-dce.checks: |

     {

       "openmetrics": {

         "init_config": {},

         "instances": [

           {

             "openmetrics_endpoint": "http://%%host%%:9000/api/monitoring/metrics",

             "metrics": [".*"],

             "auth_token": 

             {

               "reader": 

               {

                 "type": "file",

                 "path": "/etc/secret-volume/passcode"

               },

               "writer":

               {

                 "type": "header",

                 "name": "Authorization",

                 "value": "Bearer <TOKEN>"

               }

             }

           }

         ]

       }

     }