Aller au contenu

Dependencies between Dashboard and monitoring tools#

Dashboard gets some information from prometheus. This page shows the list of dependencies between Dashboard and monitoring tools.

  1. in the composant dashboard-alert-api
  • Dashboard depends on the Alert Manager service: client-alertmanager in the namespace: monitoring-client (http://client-alertmanager.monitoring-client.svc.cluster.local:9093/api/v2)
  • Dashboard depends on the Alert Manager service: app-alertmanager in the namespace: monitoring-app (http://app-alertmanager.monitoring-app.svc.cluster.local:9093/api/v2)
  • Dashboard uses the URL: https://alertmanager.ZONE_NAME.caascad.com/api/v2
  1. in the composant dashboard-metrics-api

Dashboard uses the URL: http://client-prometheus.monitoring-client.svc.cluster.local:9090/api/v1/

  1. List of promotheus requests executed by Dashboard:
  • node:node_memory_utilisation:

  • instance:node_memory_utilisation:ratio

  • request: 

    max ((1 - node_filesystem_free_bytes{mountpoint=\"/\",cc_prom_source="cluster"} / node_filesystem_size_bytes)*100) by (instance)

  • node:node_cpu_utilisation:avg1m
  • instance:node_cpu_utilisation:rate1m
  • kubelet_running_pod_count
  • request:
    sum (label_replace(container_memory_usage_bytes{cc_prom_source="cluster"} / container_spec_memory_limit_bytes{cc_prom_source="cluster"} * 100 and container_spec_memory_limit_bytes{cc_prom_source="cluster"}!= 0, \"pod_name\",\"$1\",\"existantLabel\",\"(.*)\")) by (pod_name)
  • container_fs_usage_bytes
  • container_cpu_usage_seconds_total
  • container_memory_usage_bytes
  • kube_deployment_status_replicas
  • kube_service_info
  • request: 
    max ((1 - label_replace(node_filesystem_free_bytes{mountpoint=~\"/mnt/paas/kubernetes/kubelet/plugins/kubernetes.io/flexvolume/huawei.com/fuxivol/mounts/.*\",cc_prom_source="cluster"} / node_filesystem_size_bytes{cc_prom_source="cluster"}, \"volumename\", \"$1\", \"mountpoint\", \"/mnt/paas/kubernetes/kubelet/plugins/kubernetes.io/flexvolume/huawei.com/fuxivol/mounts/(pvc-.*)\"))*100) by (volumename) * on (volumename) group_right kube_persistentvolumeclaim_info{cc_prom_source="cluster"}
  • request: 
    max ((1 - label_replace(node_filesystem_free_bytes{mountpoint=~\"/mnt/paas/kubernetes/kubelet/plugins/kubernetes.io/flexvolume/huawei.com/fuxivol/mounts/.*\"} / node_filesystem_size_bytes, \"volumename\", \"$1\", \"mountpoint\", \"/mnt/paas/kubernetes/kubelet/plugins/kubernetes.io/flexvolume/huawei.com/fuxivol/mounts/(pvc-.*)\"))*100) by (volumename) * on (volumename) group_right kube_persistentvolumeclaim_info
  • request: 
    ((sum(kube_pod_container_status_restarts_total) by (pod)) *on(pod) group_right(image_id) kube_pod_container_info)
  • request: 
    (kube_pod_start_time) * on (pod, namespace, cc_prom_source) group_right kube_pod_info
  • request: 
    (sum (kube_pod_container_status_ready) by (pod)) * on (pod) group_right() ((kube_pod_container_status_waiting_reason ==1 ) or (kube_pod_container_status_running == 1) or ( kube_pod_container_status_terminated_reason ==1))
  • kube_ingress_info
  • absent(kube_ingress_info)
  • request: 
    (kube_persistentvolume_capacity_bytes)* on (persistentvolume, namespace) group_right kube_persistentvolume_info
  • request: 
    kube_persistentvolume_status_phase == 1
  • kube_persistentvolumeclaim_info
  • kube_pod_info
  • request: 
    :caascad_cc_prom_source_kubernetes_version_1_13:
  • request: 
    :caascad_cc_prom_source_kubernetes_version_min_1_14