Enterprise Cluster Monitoring Metrics

Pods Monitoring Metrics

Namespaces to Monitor Pods

NamespacesInterpretation
ui-systemSpectro Management UI
cp-systemSystem Management UI
nats-systemMessage System
ingress-nginxIngress services
hubble-systemCore backend services
jet-systemPivot Tenant Clusters

Exceptions

The below pods are dynamically created from jobs and can be excluded from monitoring.

  • ingress-nginx-admission-patch-* [ ns: ingress-nginx ]
  • ingress-nginx-admission-create-* [ ns: ingress-nginx ]
  • packsync-* [ ns: hubble-system ]
  • cleanup-* [ ns: hubble-system ]

CPU and Memory Monitoring Metrics

Default Specifications

  • CPU: 4 vCPU
  • RAM: 8 GB RAM
  • CP Nodes: 3

Thresholds

  • CPU warn [per node ] > 70%
  • CPU alert [per node] > 80%
  • Memory Warn [per node] > 80%
  • Memory Alert [per node] > 90%

Node Monitoring Metrics

Number of Nodes: 3

Node Alerts

  • Node up
  • Node down
  • Node unreachable