通过并发，大大减少读时延 读优化 - 3 读写分离 follower 可以无限扩展，没有 raft 同步问题 读写之间无相互影响 读优化 - 4 Count 优化 基于周期性 Compact 统计，存在内存 降低时延，减轻存储压力 Watch 优化 - 1 写性能提升带来直接收益 写延迟降低，watch 延迟自然会降低 Watch 优化 - 2 纯内存态实现 无延迟损耗 超大集群，有效降低资源碎片率 落地效果 读写 QPS > 1w • 背景介绍 • 设计思路 • 性能优化 • 落地效果 • 未来演进 影响力构建 集成进入 Kubernetes 作为新型 Storage Backend 已经开源，以 TiKV 作为存储引擎 持续优化和完善系统 架构演进 • 目前所有消息严格要求有序 • 消息不重不丢、严格有序，所以写必须单点 • Kubernetes

PROVISIONER AGE 4 wordpress-sc-001 cds/nas 30m sc.yaml 内容如下 1 apiVersion: storage.k8s.io/v1 2 kind: StorageClass 3 metadata: 4 name: wordpress-sc-001 5 provisioner: nas 15 spec: 16 accessModes: 17 - ReadWriteOnce 18 resources: 19 requests: 20 storage: 20Gi 21 storageClassName: wordpress-sc-001 22 --- 23 apiVersion: v1 24 kind: PersistentVolumeClaim 30 spec: 31 accessModes: 32 - ReadWriteOnce 33 resources: 34 requests: 35 storage: 20Gi 36 storageClassName: wordpress-sc-001 创建⼀个Secret变量存放MySQL密码 2. 部署MySQL容器组 74 创建命令 1

Kubernetes满⾜了在⽣产中运⾏的应⽤程序的⼀些常⻅需求，例如： Co-locating helper processes ，促进组合应⽤程序和保留”⼀个应⽤程序的每个容器“模型 Mounting storage systems Distributing secrets Checking application health Replicating application instances Using piserver⼀个 --runtime-config 选项来禁⽤StatefulSet。 给定Pod的存储必须由 PersistentVolume Provisioner 根据请求的 storage class 提供，或由管理员预先设置。 删除/缩容StatefulSet将不会删除与StatefulSet关联的Volume。这样做是为了确保数据安全性，这通常⽐⾃动清除 StatefulSet所有相关资源更有价值。 name: pv-volume labels: type: local spec: storageClassName: manual capacity: storage: 10Gi accessModes: - ReadWriteOnce hostPath: path: "/mnt/data" --- kind: PersistentVolumeClaim

Unit December 11, 2018 2 Open Source Community Relations Engineer VMware Active in Kubernetes storage community since 2015. Chair of Kubernetes VMware SIG. GitHub: @cantbewong Staff Software Engineer tree” model, this effort spans all the touching points with the underlying infrastructure: compute, storage, network. This allows cloud providers to have independent feature and patch release cycles, learn Purpose, Projects managed, How to join The Roadmap Moving “Out of Tree” : vSphere cloud provider + storage (CSI) How to Get Information on an ongoing basis The VMware SIG How to Contribute Working Groups

n" metadata: name: "demo-01" spec: configuration: clusters: - name: "demo" Здесь нет storage Еще вернемся к этому kubectl – наше всё 1. NAMESPACE $ kubectl create namespace demo namespace/demo Persistent Storage apiVersion: "clickhouse.altinity.com/v1" kind: "ClickHouseInstallation" metadata: name: "storage" spec: defaults: templates: volumeClaimTemplate: storage-vc-template name: storage-vc-template persistentVolumeClaim: spec: accessModes: - ReadWriteOnce resources: requests: storage: 2Gi

"log-driver": "json-file", "log-opts": { "max-size": "100m" }, "storage-driver": "overlay2", "storage-opts": [ "overlay2.override_kernel_check=true" ] } systemctl restart # specific devices to use for storage can be specified for each node - name: "sdc" - name: "k8s-node3" devices: # specific devices to use for storage can be specified for each node node - name: "sdc" - name: "k8s-node4" devices: # specific devices to use for storage can be specified for each node - name: "sdc" kubectl apply -f cluster.yaml lsblk //查看磁盘

count/deployments.extensions: "2" persistentvolumeclaims: "2" #可以创建的pvc总数 requests.storage: 20Gi #可以创建的pvc存储需求存储空间总量限额 EOF # kubectl apply -f resourceQuota-testxx.yaml StorageClass的PVC只和没有指定StorageClass的PV绑定 # cat > storage-class-localdir.yaml <storage.k8s.io/v1 kind: StorageClass metadata: name: storage-class-localdir namespace: default provisioner: volumeBindingMode: WaitForFirstConsumer #等待第一个消费者使用时才绑定 （有pod使用pvc时才绑定） EOF # kubectl create -f storage-class-localdir.yaml # kubectl get sc ★创建PV（使用NFS后端） # yum install rpcbind nfs-u�ls

Architecture Storage vSphere Kubelet Datastore1 K8s Vol dataVol.vmdk K8s vSphere Cloud provider Kubernetes Worker (VM) Pod Tools, Libs, SW Redis DB K8s API vCenter Create Storage Class Create thin-disk Provisioner: vSphere Volume Diskformat: thin Name: volume-claim Storage class: thin-disk Accessmode: readwrite Storage: 2GB Podspec includes: Persistent volume claim Filesystem mount point RedisDB Comprehensive Visibility 32 UI and API Backend Advanced Analytics Engine Metrics Collection and Storage Iterate & Troubleshoot Issues Trend & Alert on Anomalies Visualize Metrics at Scale Self-Service

- volumes DevOps Operator Manage the Job environment variables image information - volumes Storage Service rbd / nfs / glisters pvc pvc pvc registry credential using secret - resources Memory user scripts using ConfigMap Job - pod template • upload files to storage service once user build completes - volumes Storage APIs user build task • build the application package init task • prepare prepare code repository sidecar build task lifecycle - preStop - volumes storage config using secret Query artifact data DevOps Operator Manage the Job CI/CD Examples - Human/Manual Task system

America December 13, 2018 2 Open Source Community Relations Engineer VMware Active in Kubernetes storage community since 2015. Chair of Kubernetes VMware SIG. GitHub: @cantbewong Application Platform interface #64021 vSphere Cloud Provider does not work when deployed across Zones with zone-local Storage #67703 24 Configuring VM affinity rules Quorum dictates design VM Fault Domain A VM VM VM VM Pro-active failure avoidance can also be engaged. Example loss of a system cooling fan, degraded storage, or can trigger automated evacuation before host failure. 26 Configuring HA restart priority Ensure