alerts/google-gke/metadata.yaml

alert_policy_templates:
-
  id: cpu-limit-utilization-all-containers
  description: "Monitors CPU limit utilization across all containers in the current project and alerts if a container's CPU limit utilization is above 90% on average for 5 minutes."
  version: 1
  related_integrations:
    - id: gke
      platform: GCP
-
  id: cpu-limit-utilization-containers-within-cluster
  description: "Monitors CPU limit utilization across all containers within a cluster and alerts if a container's CPU limit utilization is above 90% on average for 5 minutes."
  version: 1
-
  id: failedscheduling-log-event-within-cluster
  description: "Monitors log events with cluster and alerts if there is an event with reason 'FailedScheduling'."
  version: 1
-
  id: memory-limit-utilization-all-containers
  description: "Monitors memory limit utilization across all containers in the current project and alerts if a container's memory limit utilization is above 90% on average for 1 minute."
  version: 1
  related_integrations:
    - id: gke
      platform: GCP
-
  id: memory-limit-utilization-containers-within-cluster
  description: "Monitors memory limit utilization across all containers within a cluster and alerts if a container's memory limit utilization is above 90% on average for 1 minute."
  version: 1
-
  id: restarts-all-containers
  description: "Monitors all containers in the current project clusters and alerts if a container restarts more than once. (Restart count for the container is more than 1 within 5 minute window)"
  version: 1
  related_integrations:
    - id: gke
      platform: GCP
-
  id: restarts-containers-within-cluster
  description: "Monitors containers within a cluster and alerts if a container restarts more than once. (Restart count for the container is more than 1 within 5 minute window)"
  version: 1
-
  id: cpu-request-utilization-entire-workload
  description: "Alerts if the workload's CPU request utilization exceeds 90% for 5 minutes."
  version: 1
-
  id: cpu-limit-utilization-entire-workload
  description: "Alerts if the workload's CPU limit utilization exceeds 90% for 5 minutes."
  version: 1
-
  id: memory-request-utilization-entire-workload
  description: "Alerts if the workload's Memory request utilization exceeds 90% for 5 minutes."
  version: 1
-
  id: memory-limit-utilization-entire-workload
  description: "Alerts if the workload's Memory limit utilization exceeds 90% for 5 minutes."
  version: 1
-
  id: istio-server-errors-entire-workload
  description: "Alerts if the per-operation 5xx error rate exceeds 5% for 5 minutes."
  version: 1
-
  id: istio-latency-entire-workload
  description: "Alerts if the per-operation p95 latency exceeds 10 seconds for 5 minutes."
  version: 1
-
  id: istio-gmp-server-errors-entire-workload
  description: "Alerts if the 5xx error rate exceeds 5% for 5 minutes."
  version: 1
-
  id: istio-gmp-latency-entire-workload
  description: "Alerts if the p95 latency exceeds 10 seconds for 5 minutes."
  version: 1
-
  id: gke-ingress-server-errors-entire-workload
  description: "Alerts if the 5xx error rate for a URL path exceeds 5% for 5 minutes."
  version: 1
-
  id: gke-ingress-latency-entire-workload
  description: "Alerts if the p95 latency for a URL path exceeds 10 seconds for 5 minutes."
  version: 1
-
  id: nginx-ingress-server-errors-entire-workload
  description: "Alerts if the 5xx error rate for a method in a path exceeds 5% for 5 minutes."
  version: 1
-
  id: nginx-ingress-latency-entire-workload
  description: "Alerts if the p95 latency for a method in a path exceeds 10 seconds for 5 minutes."
  version: 1
-
  id: http-gmp-server-errors-entire-workload
  description: "Alerts if the 5xx error rate for a method exceeds 5% for 5 minutes."
  version: 1
-
  id: http-gmp-latency-entire-workload
  description: "Alerts if the p95 latency for a method exceeds 10 seconds for 5 minutes."
  version: 1
-
  id: grpc-server-errors-entire-workload
  description: "Alerts if the server error rate for a method exceeds 5% for 5 minutes."
  version: 1
-
  id: grpc-latency-entire-workload
  description: "Alerts if the p95 latency for a method exceeds 10 seconds for 5 minutes."
  version: 1
-
  id: utilization-container-cpu-within-workload
  description: "Alerts if a container's CPU limit utilization is above 90% on average for 5 minutes."
  version: 1
-
  id: utilization-container-memory-within-workload
  description: "Alerts if a container's memory limit utilization is above 90% on average for 1 minute."
  version: 1
-
  id: restarts-containers-within-workload
  description: "Alerts if a container restarts within a 5 minute window"
  version: 1
-
  id: cpu-limit-utilization-for-jobset
  description: "Alerts if the jobset's CPU limit utilization exceeds 90% for 5 minutes."
  version: 1
  related_integrations:
    - id: gke
      platform: GCP
-
  id: memory-limit-utilization-for-jobset
  description: "Alerts if the jobset's memory limit utilization exceeds 90% for 5 minutes."
  version: 1
  related_integrations:
    - id: gke
      platform: GCP
-
  id: memory-request-utilization-for-jobset
  description: "Alerts if the jobset's memory request utilization exceeds 90% for 5 minutes."
  version: 1
  related_integrations:
    - id: gke
      platform: GCP
-
  id: tpu-tensorcore-utilization-for-jobset
  description: "Alerts if the jobset's TPU tensorcore utilization exceeds 90% for 5 minutes."
  version: 1
  related_integrations:
    - id: gke
      platform: GCP
-
  id: jobs-failure-for-jobset
  description: "Alerts if one or more jobs fail within a jobset."
  version: 1
  related_integrations:
    - id: gke
      platform: GCP