CloudWatch Metrics

Important

Available since v1.1.0

A CloudWatch using GetMetricData can be used to obtain measurements for analysis.

Setup

You can use CloudWatch Metrics if you have used to EKS or not. This analysis is required IAM permission for cloudwatch:GetMetricData and you need to define AWS_REGION in Deployment for argo-rollouts.

EKS

If you create new cluster on EKS, you can attach cluster IAM role or attach IAM roles for service accounts. If you have already cluster on EKS, you can attach IAM roles for service accounts.

not EKS

You need to define access key and secret key.

apiVersion: v1
kind: Secret
metadata:
  name: cloudwatch-secret
type: Opaque
stringData:
  AWS_ACCESS_KEY_ID: <aws-access-key-id>
  AWS_SECRET_ACCESS_KEY: <aws-secret-access-key>
  AWS_REGION: <aws-region>

apiVersion: apps/v1
kind: Deployment
metadata:
  name: argo-rollouts
spec:
  template:
    spec:
      containers:
      - name: argo-rollouts
        env:
        - name: AWS_ACCESS_KEY_ID
          valueFrom:
            secretKeyRef:
              name: cloudwatch-secret
              key: AWS_ACCESS_KEY_ID
        - name: AWS_SECRET_ACCESS_KEY
          valueFrom:
            secretKeyRef:
              name: cloudwatch-secret
              key: AWS_SECRET_ACCESS_KEY
        - name: AWS_REGION
          valueFrom:
            secretKeyRef:
              name: cloudwatch-secret
              key: AWS_REGION

Configuration

• metricDataQueries - GetMetricData query: MetricDataQuery
• interval - optional interval, e.g. 30m, default: 5m

apiVersion: argoproj.io/v1alpha1
kind: AnalysisTemplate
metadata:
  name: success-rate
spec:
  metrics:
  - name: success-rate
    interval: 1m
    successCondition: "len(result[0].Values) >= 5 and all(result[0].Values, {# <= 0.01})"
    failureLimit: 3
    provider:
      cloudWatch:
        interval: 30m
        metricDataQueries:
        - {
            "id": "rate",
            "expression": "errors / requests"
          }
        - {
            "id": "errors",
            "metricStat": {
              "metric": {
                "namespace": "app",
                "metricName": "errors"
              },
              "period": 300,
              "stat": "Sum",
              "unit": "Count"
            },
            "returnData": false
          }
        - {
            "id": "requests",
            "metricStat": {
              "metric": {
                "namespace": "app",
                "metricName": "requests"
              },
              "period": 300,
              "stat": "Sum",
              "unit": "Count"
            },
            "returnData": false
          }

debug

You can confirm the results value in AnalysisRun.

$ kubectl get analysisrun/rollouts-name-xxxxxxxxxx-xx -o yaml
(snip)
status:
  metricResults:
  - count: 2
    failed: 1
    measurements:
    - finishedAt: "2021-09-08T17:29:14Z"
      phase: Failed
      startedAt: "2021-09-08T17:29:13Z"
      value: '[[0.0029476787030213707 0.006100422336931018 0.01020408163265306 0.007932573128408527
        0.00589622641509434 0.006339144215530904]]'
    - finishedAt: "2021-09-08T17:30:14Z"
      phase: Successful
      startedAt: "2021-09-08T17:30:14Z"
      value: '[[0.004484304932735426 0.0058374494836102376 0.006736068585425597 0.008444444444444444
        0.006859756097560976 0.0045385779122541605]]'
    name: success-rate
    phase: Running
    successful: 1
  phase: Running
  startedAt: "2021-09-08T17:29:14Z"