How deeply does Kubernetes in Action, Second Edition cover container auto-restart?

Explains restart policies, readiness changes, probe thresholds, exit states, events, and exponential back-off.

Container auto-restart

As explained in Kubernetes in Action, Second Edition

A container restart is the termination and subsequent creation or execution of a replacement instance of a container by its runtime or orchestrator.

Container auto-restart keeps a Pod’s containers running while the Pod exists. When a container’s main process terminates, the Kubelet starts it again, allowing an application to recover without special application code. Other containers in the same Pod continue running when one container fails. section 6.2.1

Watching a failed container recover

Run kubectl get pods -w, then terminate the Envoy container through its administration endpoint with curl -X POST http://localhost:9901/quitquitquit. The Pod briefly changes from Running to NotReady, shows only one ready container, and returns to Running with RESTARTS increased to 1. Kubernetes records the restart through events. section 6.2.1

A liveness probe can also trigger a restart. Kubernetes waits until the probe reaches its failureThreshold; after three consecutive failures in the example, it stops and restarts Envoy. Repeated failures produce an exponential back-off and may appear as CrashLoopBackOff. section 6.2.4

How Kubernetes in Action, Second Edition covers it

Cover of Kubernetes in Action, Second Edition

liveBook manning.com

Read Kubernetes in Action, Second Edition on liveBook Get the book at manning.com

Related concepts

container recreationimplemented by · Container recreation is the replacement of a terminated container by discarding it and creating a new container.
kubeletimplemented by · The kubelet is the Kubernetes node agent that manages Pod containers, ensures they run according to their specifications, and reports their status to the control plane.
Kubernetesimplemented by · Kubernetes is an open-source platform for orchestrating and automating the deployment, scaling, and management of containerized workloads across computing clusters.
liveness checkproduced by · A liveness check is an automated health test that determines whether a running workload remains functional or should be restarted.
volumeaddressed by · A volume is a logical storage resource that provides capacity for storing and accessing data, often by making it available to an operating system or application.
restart policyconfigured by · A restart policy is a configuration rule specifying whether and when a container or workload is automatically restarted after termination or failure.

This explainer was generated by Manning's knowledge graph from the text of Kubernetes in Action, Second Edition. The section links point to the parts of the book it draws on.