DatadogMonitoring.md

Metrics integration with Datadog

With Datadog pod annotations Datadog k8s operator will automatically collect metrics using Autodiscovery feature

Option 1: Using native Datadog integration

Pros

• No additional cost for native integrations

Cons

• Direct connection to applications is required (Ex: Datadog will connect to Postgres with user to fetch metrics)

Sample annotation for Redis

redis:
  master:
    podAnnotations:
      ad.datadoghq.com/redis.checks: |
        {
          "redisdb": {
            "init_config": {},
            "instances": [
              {
                "host": "%%host%%",
                "port":"6379",
                "password":"%%env_REDIS_PASSWORD%%"
              }
            ]
          }
        }

Option2: Using metrics exporters and collecting the same using datadog custom integration

Pros

• Compatible with most of the monitoring systems (Ex: prometheus)
• Direct connection to application is not required

Cons

• Additional cost for custom metrics in Datadog

Sample annotation for Postgres

apiVersion: postgres-operator.crunchydata.com/v1beta1
kind: PostgresCluster
metadata:
  name: simple
spec:
  ...
  instances:
    - name: instance1
      ...
      metadata:
        labels:
          tags.datadoghq.com/service: "postgrescluster-simple"
          tags.datadoghq.com/version: "14"
        annotations:
          ad.datadoghq.com/database.check_names: '["openmetrics"]'
          ad.datadoghq.com/database.init_configs: '[{}]'
          ad.datadoghq.com/database.instances: '[{"prometheus_url":"http://%%host%%:9187/metrics","namespace": "crunchy_data","metrics": ["*"]}]'