book

Hybrid Cloud Apps with OpenShift and Kubernetes

by Michael Elder, Jake Kitchener, Dr. Brad Topol

June 2021

Intermediate to advanced

271 pages

7h 19m

English

O'Reilly Media, Inc.

Read now

Unlock full access

Why We Wrote This BookWho This Book Is ForHow This Book Is OrganizedConventions Used in This BookUsing Code ExamplesO’Reilly Online LearningHow to Contact UsAcknowledgments
Kubernetes: Cloud Infrastructure for Orchestrating Containerized ApplicationsCNCF Accelerates the Growth of the Kubernetes EcosystemOpenShift: Red Hat’s Distribution of KubernetesBenefits of OpenShift for DevelopersBenefits of OpenShift for IT OperationsSummary
Kubernetes ArchitectureDeployment Options for Kubernetes and OpenShiftRed Hat’s CodeReady ContainersIBM CloudOpenShift Deployment OptionsKubernetes and OpenShift Command-Line ToolsKubernetes FundamentalsWhat’s a Pod?How Do I Describe What’s in My Pod?DeploymentsRunning the Pod and Deployment Examples in Production on OpenShiftService AccountsOpenShift EnhancementsAuthenticationProjectsApplicationsSecurity Context ConstraintsImage StreamsKubernetes and OpenShift Advanced TopicsWebhooksAdmission ControllersRole-Based Access ControlOperatorsSummary
Pod Resources and SchedulingDriving Scheduler Decisions via Resource RequestsNode Available ResourcesSchedulingPod Priority and PreemptionPost-Scheduling Pod Life CyclePod Quality of ServicePod QoS LevelsTesting Resource LimitsNode EvictionCapacity Planning and ManagementKubernetes Worker Node CapacityKubernetes Master CapacityAdmission Controller Best PracticesStandard Admission ControllersAdmission WebhooksSummary
System AvailabilityMeasuring System AvailabilityWhat Is a Highly Available System?OpenShift and Kubernetes Application and Service AvailabilityFailure ModesApplication Pod FailureWorker Node FailureWorker Zone FailureControl Plane FailureNetwork FailureSummary
HelmKustomizeGeneratorsCompositionPatchesOverlaysDirect Deploy of Kustomize-Generated Resource FilesGitOpsRazeeArgo CDTektonTasksPipelinesOpenShift PipelinesOpen Cluster Management AppsSummary
Why Multicluster?Use Case: Using Multiple Clusters to Provide Regional Availability for Your ApplicationsUse Case: Using Multiple Clusters for Multiple TenantsUse Case: Supporting Far-Edge Use Cases Where Clusters Do Not Run in Traditional Datacenters or CloudsArchitectural CharacteristicsProvisioning Across CloudsUser-Managed OpenShiftUpgrading Your Clusters to the Latest Version of KubernetesSummary of Multicloud Cluster ProvisioningOpenShift as a ServiceKubernetes as a ServiceOperating System Currency for NodesSummary
Configuring Your Cluster with OperatorsUnderstanding OperatorsExample: Container Security OperatorUsing Cluster Operators to Manage OpenShiftExample: Configuring the Authentication OperatorOpenShift htpasswd Identity ProviderPolicy and Compliance Across Multiple ClustersPolicy Example: Federate a Project Across Your FleetPlacementRules to Assign Content to ManagedClustersPolicy Example: Managing etcd Encryption Within ManagedClustersPolicy Example: Managing RBAC Within ManagedClustersPolicy Example: Managing IdPs Within ManagedClustersPolicy Example: Managing Upgrades with Policy Across ManagedClustersSummary
Failure Is InevitableMulticluster Load BalancingAutomating Without OperatorsDeploying the Example ApplicationConfigure Your Hub ClusterProvision an OpenShift Cluster to Host the Open Cluster Management HubConfigure the Open Cluster Management HubProvision Two or More Clusters to Host the ApplicationDeploy Ansible Tower and the Ansible Resource OperatorConfigure Projects for ServiceNow and F5 DNS Load BalancerConfigure the toweraccess Secret and Create the Ansible Tower TokenDeploy the pacman-app Example to Your ClusterSummary

Increased Migration of Legacy Enterprise Applications to Cloud Native ApplicationsIncreased Adoption of Kubernetes for High Performance ComputingKubernetes and OpenShift Will Become the De Facto Platforms for Machine Learning and Deep LearningOpen Cloud Marketplaces Will Accelerate Adoption of Cloud Native ApplicationsOpenShift Will Be the Platform for Enterprise MulticloudRecommended ResourcesIBM Developer WebsiteLearn OpenShiftKubernetes WebsiteKubernetes IBM Cloud Provider Special Interest GroupKubernetes Contributor Experience Special Interest GroupSummary

Content preview from Hybrid Cloud Apps with OpenShift and Kubernetes

Chapter 8. Working Example of Multicluster Application Delivery

Let’s take a look at a simple working example of a web application with a backing datastore. For our purposes, we will deploy an app that mimics the game PAC-MAN game from Atari. A user will interact with a dynamic frontend that stores information in a backing MongoDB.

We will deploy this application across multiple distinct clusters using the techniques discussed in Chapter 5. Each of the clusters will be provisioned from a hub running Open Cluster Management, as discussed in Chapter 6. Further, we will configure an external load balancer provided by a GSLB service hosted by F5. Incoming user requests will be routed from a global domain into one of the specific clusters. If any one of the clusters or the application experiences a problem, then user requests will no longer be routed to that cluster. We are going to go a little further and demonstrate how to integrate off-cluster resources like an F5 DNS Load Balancer Cloud Service, and we will integrate a ServiceNow change ticket into the example.

In Figure 8-1, we see our PAC-MAN application running on two OpenShift clusters with a load balancer routing traffic to app instances on either cluster. We can see the hub cluster with various resources that are helping to manage the overall system. Through the rest of this chapter, we will explain what these various parts are doing and provide a walk-through that you can do on your own to experiment with all of the moving ...