Overview/Description
Monitoring the current state of system components is one of the most efficient ways to identify problems and bottlenecks in a production environment. Administrators working with Kubernetes need to know how to monitor at multiple levels, including at the node, cluster, and Pod levels. In this course, you'll examine the vital Kubernetes performance metrics to monitor, what to measure in Kubernetes clusters, nodes, and Pods, and best practices for monitoring these. You'll also identify the differences between logging and monitoring. Moving on, you'll write configuration files to create and start node-problem-detector. You'll then specify Pod configuration files to ensure kubelet performs liveness probes, HTTP GET requests are sent to the server running in the container, and TCP liveness probe readiness is defined. This course is part of a series that aligns with the Certified Kubernetes Administrator exam's objectives and can be used in preparation for it.
describe the goal of monitoring Kubernetes clusters and the differences between logging and monitoring
list the key elements that need to be monitored in a Kubernetes cluster and recall the differences between whitebox and blackbox monitoring
recognize the Kubernetes cluster metrics that help reveal the resource utilization of the entire cluster and outline how to collect metrics from Kubernetes clusters and export them to external endpoints
create a configuration file and write configurations to create and start node-problem-detector, which monitors and detects node problems
update the Kernel Monitor by editing the Conditions field in the config/kernel-monitor.json configuration file using the new condition definition
list the various states of an individual Kubernetes cluster and its master and worker nodes
describe the best practices that should be employed during Kubernetes monitoring
build a configuration file that can be used to create a Pod running a container and specify a configuration that ensures the kubelet performs a liveness probe every 10 seconds and waits for 10 seconds before performing the first probe
build a configuration file that can be used to create a Pod and specify a configuration that ensures the kubelet sends an HTTP GET request to the server running in the container and listening at the port 8080
build a configuration file that can be used to create a Pod and specify a configuration that defines the readiness of TCP liveness probes
list the key Kubernetes performance metrics that need to be monitored to get data about the count, health, and availability of various Kubernetes objects
query the Metrics API to retrieve the current metrics from any node or Pod along with Kubernetes node CPU and memory usage