Goldilocks

Goldilocks (opens new window) watches your Kubernetes deployments and makes recommendations for resource requests and limits based on actual usage.

Goldilocks utilizes the vertical-pod-autoscaler (opens new window) in recommendation mode to extract suggested CPU and memory limits/requests.

Requirements

Make sure you have metrics-server (opens new window) installed in your cluster.

You'll also need vertical-pod-autoscaler (opens new window). Goldilocks will install this for you by default, but if you're managing your own VPA installation, you can --set goldilocks.installVPA=false.

Remediation

Goldilocks Action Items will suggest particular amounts for CPU and memory limits/requests. You can copy/paste these recommendations into your workload configuration.

If Goldilocks is making recommendations for a third-party application, you can likely set its requests/limits using its Helm chart or by editing its installation YAML by hand.

Sample Report

Goldilocks reports contain information about resource usage for each Deployment.

{
    "deployments": [
        {
            "containers": [
                {
                    "containerName": "fairwinds-insights",
                    "limits": {
                        "cpu": "500m",
                        "memory": "1Gi"
                    },
                    "lowerBound": {
                        "cpu": "25m",
                        "memory": "262144k"
                    },
                    "requests": {
                        "cpu": "500m",
                        "memory": "1Gi"
                    },
                    "target": {
                        "cpu": "1000m",
                        "memory": "3Gi"
                    },
                    "uncappedTarget": {
                        "cpu": "2000m",
                        "memory": "3Gi"
                    },
                    "upperBound": {
                        "cpu": "2000m",
                        "memory": "3Gi"
                    }
                }
            ],
            "deploymentName": "my-app",
            "namespace": "my-app"
        }
    ]
}