best-practice-karmada-istio-improve-workload-traffic-resilience-of-production-distributed-cloud-huan-daepkarmadareistiopan-chang-zha-du-zhe-rehao-chaomeng-zhang-huawei.pdf

1、Best practice:Karmada&Istio improve workload&traffic resilience of production distributed About MeArchitect of HUAWEI Distributed Cloud Native,with 9 years cloud computing related design and developing experience in HUAWEI Cloud,including service mesh,Kubernetes,micro service,cloud service catalog,b

2、ig data,APM,cloud computing reliability and DevOps.He is Istio community member,an experienced speaker of KubeCon,IstioCon,ServiceMeshCon,author of books“Cloud Native Service Mesh Istio”(云原生服务网格 Istio)and“Istio:the Definitive Guide”(Istio权威指南)AgendaBackend:Resilience challenges and Kubernetes&Istios

3、 solutionDistributed Cloud resilience improvement and new challengesPracticeKarmada improve workload resilience of multi-cloud envIstio improve traffic resilience of multi-cloud envKarmada and Istio total multi-cloud application resilience Resilience Resilience：the ability of a system or organizatio

4、n to respond to or recover readily from a crisis,disruptive process,etc.Kubernetes Resilience(1)DeploymentReplica SetPodPodNodePodDeploymentReplica SetPodPodPodNodeNodePodPodPodServiceEndpointsKubernetes Resilience(2)ReplicaSet desired number of pods,automatically scaling ServiceLoad balancing betwe

5、en podsDeploymentEnsuring desired state of applicationLoggingCICDSecretReadinessLivenessRolling updateHPALoadbalanceConfigMapRBACFault toleranceAnti-affinityIstio Resilience(1)IstioConsumerEnvoyProducerEnvoyProducerProducerEnvoyIstio Resilience(2)Rolling UpdateAnti-affinityRedundancyAuto scalingCirc

6、uit breakerTimeoutRetryRate limitFault InjectionTestingInfrastructureApplicationOpsTracingAccess logLocality CanaryMetricHealth CheckLoadbalanceKubernetes Resilience ChallengesPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodPodDistributed Clouds SolutionPublic CloudPremiseOther