Expose HiveMQ JMX metrics for DataDog

This artical explains how to expose HiveMQ JMX metrics in DataDog using Kubernetes

\uD83D\uDCD8 Pre-requisites

1. Helm version v3+

2. Running Kubernetes cluster version 1.18.0 or higher

3. kubectl lastest version

\uD83D\uDCD8 Instructions

1. Configure DataDog agent using the Helm chart

   1. Add Datadog helm repository

      1. helm repo add datadog https://helm.datadoghq.com
         helm repo update

   2. Create datadog-values.yaml. For full values check here
      Please refer to following the example of datadog-values.yaml

      1. logLevel: DEBUG
         targetSystem: "linux"
         datadog:
           apiKey: <your api key from your account>
           
           #Logs configurations
           logs:
             enabled: true
             containerCollectAll: true
             containerCollectUsingFiles: true
           
           #Plesse configure this setting based on your region
           site: datadoghq.eu
           
           
           tags: []
           # datadog.kubelet.tlsVerify should be `false` on kind and minikube
           # to establish communication with the kubelet
           kubelet:
             tlsVerify: false
           
           #Following settings are required to collect the Kubernetes events with the Cluster Agent  
           collectEvents: true
           clusterAgent:
             rbac:
               create: true
           
         
         # setting to read JMX metrics from HiveMQ
         agents:
           image:
             tagSuffix: "jmx"

   3. install Datadog agent using Helm

      1. kubectl create ns <namespace>
         helm install -f datadog-values.yaml <release name> datadog/datadog 

   4. Verify installation and check the agent pod’s status

      1. kubectl get pods

      2. Also, check the agent’s log’s to confirm.

2. Install HiveMQ using Operator

   1. Configure hivemq-values.yaml based on your requirement and then add podAnnotations for Datadog to auto-discover HiveMQ metrics and logs. Check here for details.
      Please find below a minimalist config example of hivemq-values.yaml

      1. hivemq:
           image: hivemq/hivemq4:k8s-4.7.5
           imagePullPolicy: Always
           cpu: 8
           memory: 8Gi
           nodeCount: "3"
         
         
           podAnnotations:
             ad.datadoghq.com/hivemq.check_names: '[\"hivemq\"]'
             ad.datadoghq.com/hivemq.init_configs: '[{\"is_jmx\": true, \"collect_default_metrics\": true}]'
             ad.datadoghq.com/hivemq.instances: '[{\"host\": \"%%host%%\",\"port\":\"9010\"}]'
             ad.datadoghq.com/hivemq.logs: '[{\"source\":\"hivemq\",\"service\":\"hivemq-hivemq-1778-mqtt\"}]'
           
           ports:
             - name: "mqtt"
               port: 1883
               expose: true
               patch:
                - '[{"op":"add","path":"/spec/type","value":"LoadBalancer"}]'
             - name: "cc"
               port: 8080
               expose: true
               patch:
                 - '[{"op":"add","path":"/spec/sessionAffinity","value":"ClientIP"}]'
             
         operator:
           admissionWebhooks:
             # Enable the admission hook
             enabled: false
         
         monitoring:
           enabled: false
           dedicated: false
         
         

   2. Install hivemq using the following command

      1. helm upgrade --install -f hivemq-values.yaml <release name> hivemq/hivemq-operator

   3. verify installation

      1. kubectl get pods

   4. Login to your Datadog account and click “Metrics” =>” Explorer” from the left navigation links. You will be redirected to the “Graph your data” page. By default, you will see “system.cpu.user“ metrics and its data.
      Type hivemq in the search box and you should be able to see the list of available metrics.

   5. 

      Check HiveMQ dashboard
      

      

3. Uninstall DataDog and HiveMQ

   1. helm list
      helm uninstall <datadog release name>
      helm uninstall <hivemq release name>
      kubectl delete ns <namespce>