book

Managing Kubernetes

by Brendan Burns, Craig Tracey

November 2018

Intermediate to advanced

185 pages

4h 39m

English

O'Reilly Media, Inc.

Read now

Unlock full access

Who should read This BookWhy we wrote This BookNavigating This BookConventions Used in This BookUsing Code ExamplesO’Reilly SafariHow to Contact UsAcknowledgments
How the Cluster OperatesAdjust, Secure, and Tune the ClusterResponding When Things Go WrongExtending the System with New and Custom FunctionalitySummary
ContainersContainer OrchestrationThe Kubernetes APIBasic Objects: Pods, ReplicaSets, and ServicesOrganizing Your Cluster with Namespaces, Labels, and AnnotationsAdvanced Concepts: Deployments, Ingress, and StatefulSetsBatch Workloads: Job and ScheduledJobCluster Agents and Utilities: DaemonSetsSummary
ConceptsDeclarative ConfigurationReconciliation or ControllersImplicit or Dynamic GroupingStructureUnix Philosophy of Many ComponentsAPI-Driven InteractionsComponentsHead Node ComponentsComponents On All NodesScheduled ComponentsSummary
Basic Characteristics for ManageabilityPieces of the API ServerAPI ManagementAPI PathsAPI DiscoveryOpenAPI Spec ServingAPI TranslationRequest ManagementTypes of RequestsLife of a RequestAPI Server InternalsCRD Control LoopDebugging the API ServerBasic LogsAudit LogsActivating Additional LogsDebugging kubectl RequestsSummary
An Overview of SchedulingScheduling ProcessPredicatesPrioritiesHigh-Level AlgorithmConflictsControlling Scheduling with Labels, Affinity, Taints, and TolerationsNode SelectorsNode AffinityTaints and TolerationsSummary
kubeadmRequirementskubeletInstalling the Control Planekubeadm ConfigurationPreflight ChecksCertificatesetcdkubeconfigTaintsInstalling Worker NodesAdd-OnsPhasesHigh AvailabilityUpgradesSummary
UsersAuthenticationkubeconfigService AccountsSummary
RESTAuthorizationRole-Based Access ControlRole and ClusterRoleRoleBinding and ClusterRoleBindingTesting AuthorizationSummary
ConfigurationCommon ControllersPodSecurityPoliciesResourceQuotaLimitRangeDynamic Admission ControllersValidating Admission ControllersMutating Admission ControllersSummary

Container Network InterfaceChoosing a Plug-inkube-proxyService DiscoveryDNSEnvironment VariablesNetwork PolicyService MeshSummary
Goals for MonitoringDifferences Between Logging and MonitoringBuilding a Monitoring StackGetting Data from Your Cluster and ApplicationsAggregating Metrics and Logs from Multiple SourcesStoring Data for Retrieval and QueryingVisualizing and Interacting with Your DataWhat to Monitor?Monitoring MachinesMonitoring KubernetesMonitoring ApplicationsBlackbox MonitoringStreaming LogsAlertingSummary
High AvailabilityStateApplication DataPersistent VolumesLocal DataWorker NodesetcdArkSummary
Kubernetes Extension PointsCluster DaemonsUse Cases for Cluster DaemonsInstalling a Cluster DaemonOperational Considerations for Cluster DaemonsHands-On: Example of Creating a Cluster DaemonCluster AssistantsUse Cases for Cluster AssistantsInstalling a Cluster AssistantOperational Considerations for Cluster AssistantsHands-On: Example of Cluster AssistantsExtending the Life Cycle of the API ServerUse Cases for Extending the API Life CycleInstalling API Life Cycle ExtensionsOperational Considerations for Life Cycle ExtensionsHands-On: Example of Life Cycle ExtensionsAdding Custom APIs to KubernetesUse Cases for Adding New APIsCustom Resource Definitions and Aggregated API ServersArchitecture for Custom Resource DefinitionsInstalling Custom Resource DefinitionsOperational Considerations for Custom ResourcesSummary

Content preview from Managing Kubernetes

Chapter 13. Extending Kubernetes

Kubernetes has a rich API that provides much of the functionality that you might need to build and operate a distributed system. However, the API is purposefully generic, aimed at the 80% use cases. Taking advantage of the rich ecosystem of add-ons and extensions that exist for Kubernetes can add significant new functionality and enable new experiences for users of your cluster. You may even choose to implement your own custom add-ons and extensions that are suited to the particular needs of your company or environment.

Kubernetes Extension Points

There are a number of different ways to extend a Kubernetes cluster, and each offers a different set of capabilities and additional operational complexity. The following sections describe these various extension points in detail and provide insight into how they can extend the functionality of a cluster and into the additional operational requirements of these extensions.

The four types of extensibility are:

Cluster daemons for automation
Cluster assistants for extended functionality
Extending the life cycle of the API server
Adding more APIs

The truth, of course, about some of these classifications is that they are somewhat arbitrary, and there are different extensions that can combine multiple kinds of extensibility to provide additional functionality for a cluster. The categories described here are intended to help guide your discussion and planning for extending a Kubernetes cluster. They ...