Kubernetes performance monitoring: cluster, pods & more

• Get a cluster-level overview of pods, deployments, CPU, and memory allocation
• Monitor the master node and all worker nodes in a Kube cluster
• For each node, monitor pod allocation image usage, CPU, and memory allocation and utilization
• Track all namespaces associated with the cluster and monitor the activation status, age and pod, and service association for each namespace

The Kubernetes control plane is one of the key elements to monitor in the Kubernetes environment. The control plane runs on the master node and is composed of 4 critical services: API server, controller manager, etcd, and scheduler.

eG Enterprise monitors the status of all these critical services and alerts you to service failures. Also drill down to see specific messages and notifications for each service to help understand failures.

Monitor all cluster events from one place: container, pod, node, deployment, DaemonSet, image, image manager, pod worker, jobs, scheduler, load balancer, and so on.

• For each pod in a namespace, track pod readiness and schedule, and CPU and memory allocation and utilization
• Understand resource usage patterns and determine how much additional workload can be deployed per node
• For each DaemonSet in a namespace, monitor DaemonSet availability on nodes, missed scheduled, details of pods, and nodes running the DaemonSet
• Monitor application services associated to each namespace and track their status, service type age, and which pods are connected to each service
• Get alerted to failed jobs, failed pods, and longest running jobs

In context of Kubernetes operations, eG Enterprise provides visibility of the Docker hosts and containers provisioned on them automatically.

• Track the container lifecycle state from creation to deletion; know which containers are running, stopped, and paused
• Monitor Docker services and events, and get alerted to any anomalies
• Measure CPU, memory, and disk utilization of every Docker container, and get alerted to high usage levels
• Obtain granular visibility into the Host's resource availability and utilization: disk space, disk activity, I/O waits, memory, swap, and inodes
• Monitor the images mapped and not mapped to containers

• How is the health and availability of the Kubernetes Deployment?
• Are all Master Node services running as expected?
• Are there any failed Container images while pulling from the image repository?
• What is the overall view of the Cluster in terms of Pod allocation, Deployment status, CPU, and memory allocation?
• Are DaemonSets unavailable on any Nodes?
• What is the status of Namespaces and Services? Are they properly mapped to their corresponding Pods?