K8S与Ceph RBD集成-多主与主从数据库示例
参考文章:
https://ieevee.com/tech/2018/05/16/k8s-rbd.html
https://zhangchenchen.github.io/2017/11/17/kubernetes-integrate-with-ceph/
https://docs.openshift.com/container-platform/3.5/install_config/storage_examples/ceph_rbd_dynamic_example.html
https://jimmysong.io/kubernetes-handbook/practice/using-ceph-for-persistent-storage.html
感谢以上作者提供的技术参考,这里我加以整理,分别实现了多主数据库集群和主从数据库结合Ceph RDB的实现方式。以下配置只为测试使用,不能做为生产配置。
K8S中存储的分类
在K8S的持久化存储中主要有以下几种分类:
volume: 就是直接挂载在pod上的组件,k8s中所有的其他存储组件都是通过volume来跟pod直接联系的。volume有个type属性,type决定了挂载的存储是什么,常见的比如:emptyDir,hostPath,nfs,rbd,以及下文要说的persistentVolumeClaim等。跟docker里面的volume概念不同的是,docker里的volume的生命周期是跟docker紧紧绑在一起的。这里根据type的不同,生命周期也不同,比如emptyDir类型的就是跟docker一样,pod挂掉,对应的volume也就消失了,而其他类型的都是永久存储。详细介绍可以参考Volumes
Persistent Volumes:顾名思义,这个组件就是用来支持永久存储的,Persistent Volumes组件会抽象后端存储的提供者(也就是上文中volume中的type)和消费者(即具体哪个pod使用)。该组件提供了PersistentVolume和PersistentVolumeClaim两个概念来抽象上述两者。一个PersistentVolume(简称PV)就是后端存储提供的一块存储空间,具体到ceph rbd中就是一个image,一个PersistentVolumeClaim(简称PVC)可以看做是用户对PV的请求,PVC会跟某个PV绑定,然后某个具体pod会在volume 中挂载PVC,就挂载了对应的PV。
- Dynamic Volume Provisioning: 动态volume发现,比如上面的Persistent Volumes,我们必须先要创建一个存储块,比如一个ceph中的image,然后将该image绑定PV,才能使用。这种静态的绑定模式太僵硬,每次申请存储都要向存储提供者索要一份存储快。Dynamic Volume Provisioning就是解决这个问题的。它引入了StorageClass这个概念,StorageClass抽象了存储提供者,只需在PVC中指定StorageClass,然后说明要多大的存储就可以了,存储提供者会根据需求动态创建所需存储快。甚至于,我们可以指定一个默认StorageClass,这样,只需创建PVC就可以了。
配置初始化环境
- 已经有一个k8s集群
- 已经有一个Ceph 集群
所有节点安装ceph-common
添加ceph的yum源:
[Ceph]name=Ceph packages for $basearchbaseurl=https://mirrors.aliyun.com/ceph/rpm-mimic/el7/$basearchenabled=1gpgcheck=1type=rpm-mdgpgkey=https://download.ceph.com/keys/release.asc[Ceph-noarch]name=Ceph noarch packagesbaseurl=https://mirrors.aliyun.com/ceph/rpm-mimic/el7/noarchenabled=1gpgcheck=1type=rpm-mdgpgkey=https://download.ceph.com/keys/release.asc[ceph-source]name=Ceph source packagesbaseurl=https://mirrors.aliyun.com/ceph/rpm-mimic/el7/SRPMSenabled=1gpgcheck=1type=rpm-mdgpgkey=https://download.ceph.com/keys/release.asc
安装ceph-common:
yum install ceph-common -y
如果安装过程出现依赖报错,可以通过如下方式解决:
yum install -y yum-utils && \yum-config-manager --add-repo https://dl.fedoraproject.org/pub/epel/7/x86_64/ && \yum install --nogpgcheck -y epel-release && \rpm --import /etc/pki/rpm-gpg/RPM-GPG-KEY-EPEL-7 && \rm -f /etc/yum.repos.d/dl.fedoraproject.org*yum -y install ceph-common
配置ceph配置文件
将ceph配置文件拷贝到各个k8s的node节点
[root@ceph-1 ~]# scp /etc/ceph k8s-node:/etc/
测试volume
通过使用一个简单的volume,测试集群环境是否正常,在实际的应用中,需要永久保存的数据不能使用volume的方式。
在Ceph集群中创建images
创建新的镜像时,需要禁用某些不支持的属性:
rbd create foobar -s 1024 -p k8s rbd feature disable k8s/foobar object-map fast-diff deep-flatten
查看镜像信息:
# rbd info k8s/foobarrbd image 'foobar': size 1 GiB in 256 objects order 22 (4 MiB objects) id: ad9b6b8b4567 block_name_prefix: rbd_data.ad9b6b8b4567 format: 2 features: layering, exclusive-lock op_features: flags: create_timestamp: Tue Apr 23 17:37:39 2019
使用POD直接挂载volume
这里指定了ceph的 admin.keyring文件作为认证密钥:
# cat test.yaml apiVersion: v1kind: Podmetadata: name: rbdspec: containers: - image: nginx name: rbd-rw volumeMounts: - name: rbdpd mountPath: /mnt volumes: - name: rbdpd rbd: monitors: - '192.168.20.41:6789' pool: k8s image: foobar fsType: xfs readOnly: false user: admin keyring: /etc/ceph/ceph.client.admin.keyring
使用PV和PVC
如果需要永久保存数据(当pod删除后数据不会丢失),我们需要使用PV(PersistentVolume),和PVC(PersistentVolumeClaim)的方式。
在Ceph集群中创建images
rbd create -s 1024 k8s/pvrbd feature disable k8s/pv object-map fast-diff deep-flatten
查看镜像信息:
# rbd info k8s/pvrbd image 'pv': size 1 GiB in 256 objects order 22 (4 MiB objects) id: adaa6b8b4567 block_name_prefix: rbd_data.adaa6b8b4567 format: 2 features: layering, exclusive-lock op_features: flags: create_timestamp: Tue Apr 23 19:09:58 2019
创建一个secret
- 生成一个加密的key
grep key /etc/ceph/ceph.client.admin.keyring |awk '{printf "%s", $NF}'|base64
- 将生成的key创建一个Secret
apiVersion: v1kind: Secretmetadata: name: ceph-secrettype: "kubernetes.io/rbd" data: key: QVFBbk1MaGNBV2laSGhBQUVOQThRWGZyQ3haRkJDNlJaWTNJY1E9PQ==---
创建PV和PVC文件
# cat ceph-rbd-pv.yaml apiVersion: v1kind: PersistentVolumemetadata: name: ceph-rbd-pvspec: capacity: storage: 1Gi accessModes: - ReadWriteOnce rbd: monitors: - '192.168.20.41:6789' pool: k8s image: pv user: admin secretRef: name: ceph-secret fsType: xfs readOnly: false persistentVolumeReclaimPolicy: Recycle# cat ceph-rbd-pvc.yaml apiVersion: v1kind: PersistentVolumeClaimmetadata: name: ceph-rbd-pv-claimspec: accessModes: - ReadWriteOnce resources: requests: storage: 1Gi
创建POD
# cat test3-pvc.yaml apiVersion: v1kind: Podmetadata: name: rbd-nginxspec: containers: - image: nginx name: rbd-rw volumeMounts: - name: rbd-pvc mountPath: /mnt volumes: - name: rbd-pvc persistentVolumeClaim: claimName: ceph-rbd-pv-claim
使用StorageClass
Storage Class的作用
简单来说,storage配置了要访问ceph RBD的IP/Port、用户名、keyring、pool,等信息,我们不需要提前创建image;当用户创建一个PVC时,k8s查找是否有符合PVC请求的storage class类型,如果有,则依次执行如下操作:
- 到ceph集群上创建image
- 创建一个PV,名字为pvc-xx-xxx-xxx,大小pvc请求的storage。
- 将上面的PV与PVC绑定,格式化后挂到容器中
通过这种方式管理员只要创建好storage class就行了,后面的事情用户自己就可以搞定了。如果想要防止资源被耗尽,可以设置一下Resource Quota。
当pod需要一个卷时,直接通过PVC声明,就可以根据需求创建符合要求的持久卷。
创建storage class
# cat storageclass.yaml apiVersion: storage.k8s.io/v1kind: StorageClassmetadata: name: fastprovisioner: kubernetes.io/rbdparameters: monitors: 192.168.20.41:6789 adminId: admin adminSecretName: ceph-secret pool: k8s userId: admin userSecretName: ceph-secret fsType: xfs imageFormat: "2" imageFeatures: "layering"
创建PVC
RBD只支持 ReadWriteOnce 和 ReadOnlyAll,不支持ReadWriteAll。注意这两者的区别点是,不同nodes之间是否可以同时挂载。同一个node上,即使是ReadWriteOnce,也可以同时挂载到2个容器上的。
创建应用的时候,需要同时创建 pv和pod,二者通过storageClassName关联。pvc中需要指定其storageClassName为上面创建的sc的name(即fast)。
# cat pvc.yaml kind: PersistentVolumeClaimapiVersion: v1metadata: name: rbd-pvc-pod-pvcspec: accessModes: - ReadWriteOnce volumeMode: Filesystem resources: requests: storage: 1Gi storageClassName: fast
创建pod
# cat pod.yaml apiVersion: v1kind: Podmetadata: labels: test: rbd-pvc-pod name: ceph-rbd-sc-pod1spec: containers: - name: ceph-rbd-sc-nginx image: nginx volumeMounts: - name: ceph-rbd-vol1 mountPath: /mnt readOnly: false volumes: - name: ceph-rbd-vol1 persistentVolumeClaim: claimName: rbd-pvc-pod-pvc
补充
在使用Storage Class时,除了使用PVC的方式声明要使用的持久卷,还可通过创建一个volumeClaimTemplates进行声明创建(StatefulSets中的存储设置),如果涉及到多个副本,可以使用StatefulSets配置:
apiVersion: apps/v1kind: StatefulSetmetadata: name: nginxspec: selector: matchLabels: app: nginx serviceName: "nginx" replicas: 3 template: metadata: labels: app: nginx spec: terminationGracePeriodSeconds: 10 containers: - name: nginx image: nginx volumeMounts: - name: www mountPath: /usr/share/nginx/html volumeClaimTemplates: - metadata: name: www spec: accessModes: [ "ReadWriteOnce" ] storageClassName: "fast" resources: requests: storage: 1Gi
但注意不要用Deployment。因为,如果Deployment的副本数是1,那么还是可以用的,跟Pod一致;但如果副本数 >1 ,此时创建deployment后会发现,只启动了1个Pod,其他Pod都在ContainerCreating状态。过一段时间describe pod可以看到,等volume等很久都没等到。
示例一:创建一个mysql-galera集群(多主)
官方文档:https://kubernetes.io/docs/tasks/run-application/run-replicated-stateful-application/
statefulset简介
statefulset(1.5之前叫做petset),statefulset与deployment,replicasets是一个级别的。不过Deployments和ReplicaSets是为无状态服务而设计。statefulset则是为了解决有状态服务的问题。它的应用场景如下:
- 稳定的持久化存储,即Pod重新调度后还是能访问到相同的持久化数据,基于PVC来实现
- 稳定的网络标志,即Pod重新调度后其PodName和HostName不变,基于Headless Service(即没有Cluster IP的Service)来实现。
- 有序部署,有序扩展,即Pod是有顺序的,在部署或者扩展的时候要依据定义的顺序依次依次进行(即从0到N-1,在下一个Pod运行之- 前所有之前的Pod必须都是Running和Ready状态),基于init containers来实现。
- 有序收缩,有序删除(即从N-1到0)。
由应用场景可知,statefuleset特别适合mqsql,redis等数据库集群。相应的,一个statefuleset有以下三个部分:
- 用于定义网络标志(DNS domain)的HeadlessService,参考文档 )
- 用于创建PersistentVolumes的volumeClaimTemplates
- 定义具体应用的StatefulSet
1. 生成并创建ceph secret
如果k8s集群中已经创建了ceph 的secret可以跳过此步
生成一个加密的key
grep key /etc/ceph/ceph.client.admin.keyring |awk '{printf "%s", $NF}'|base64
将生成的key创建一个Secret
apiVersion: v1kind: Secretmetadata: name: ceph-secret namespace: galeratype: "kubernetes.io/rbd" data: key: QVFBbk1MaGNBV2laSGhBQUVOQThRWGZyQ3haRkJDNlJaWTNJY1E9PQ==---
2. 创建StorageClass
# cat storageclass.yaml apiVersion: storage.k8s.io/v1kind: StorageClassmetadata: name: fastprovisioner: kubernetes.io/rbdparameters: monitors: 192.168.20.41:6789,192.168.20.42:6789,192.168.20.43:6789 adminId: admin adminSecretName: ceph-secret pool: k8s userId: admin userSecretName: ceph-secret fsType: xfs imageFormat: "2" imageFeatures: "layering"
3. 创建headless Service
galera-service.yaml
apiVersion: v1kind: Servicemetadata: annotations: service.alpha.kubernetes.io/tolerate-unready-endpoints: "true" name: galera namespace: galera labels: app: mysqlspec: ports: - port: 3306 name: mysql # *.galear.default.svc.cluster.local clusterIP: None selector: app: mysql
4. 创建statefulset
这里使用V1版本的StatefulSet,和之前的版本相比,v1版本是当前的稳定版本,同时与之前的beta版的区别是v1版本需要添加spec.selector.matchLabels的参数,此参数需要与spec.template.metadata.labels保持一致。
apiVersion: apps/v1kind: StatefulSetmetadata: name: mysql namespace: galeraspec: selector: matchLabels: app: mysql serviceName: "galera" replicas: 3 template: metadata: labels: app: mysql spec: initContainers: - name: install image: mirrorgooglecontainers/galera-install:0.1 imagePullPolicy: Always args: - "--work-dir=/work-dir" volumeMounts: - name: workdir mountPath: "/work-dir" - name: config mountPath: "/etc/mysql" - name: bootstrap image: debian:jessie command: - "/work-dir/peer-finder" args: - -on-start="/work-dir/on-start.sh" - "-service=galera" env: - name: POD_NAMESPACE valueFrom: fieldRef: apiVersion: v1 fieldPath: metadata.namespace volumeMounts: - name: workdir mountPath: "/work-dir" - name: config mountPath: "/etc/mysql" containers: - name: mysql image: mirrorgooglecontainers/mysql-galera:e2e ports: - containerPort: 3306 name: mysql - containerPort: 4444 name: sst - containerPort: 4567 name: replication - containerPort: 4568 name: ist args: - --defaults-file=/etc/mysql/my-galera.cnf - --user=root readinessProbe: # TODO: If docker exec is buggy just use gcr.io/google_containers/mysql-healthz:1.0 exec: command: - sh - -c - "mysql -u root -e 'show databases;'" initialDelaySeconds: 15 timeoutSeconds: 5 successThreshold: 2 volumeMounts: - name: datadir mountPath: /var/lib/ - name: config mountPath: /etc/mysql volumes: - name: config emptyDir: {} - name: workdir emptyDir: {} volumeClaimTemplates: - metadata: name: datadir annotations: volume.beta.kubernetes.io/storage-class: "fast" spec: accessModes: [ "ReadWriteOnce" ] resources: requests: storage: 1Gi
5. 检查pod
查看pod状态已经正常
[root@master-1 ~]# kubectl get pod -n galera NAME READY STATUS RESTARTS AGEmysql-0 1/1 Running 0 48mmysql-1 1/1 Running 0 43mmysql-2 1/1 Running 0 38m
数据库集群建立:
[root@master-1 ~]# kubectl exec mysql-1 -n galera -- mysql -uroot -e 'show status like "wsrep_cluster_size";'Variable_name Valuewsrep_cluster_size 3
查看pv绑定:
[root@master-1 mysql-cluster]# kubectl get pvc -l app=mysql -n galeraNAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGEdatadir-mysql-0 Bound pvc-6e5a1c45-666b-11e9-ad20-000c29016590 1Gi RWO fast 3d20hdatadir-mysql-1 Bound pvc-25683cfd-666c-11e9-ad20-000c29016590 1Gi RWO fast 3d20hdatadir-mysql-2 Bound pvc-c024b422-666c-11e9-ad20-000c29016590 1Gi RWO fast 3d20h
测试数据库:
kubectl exec mysql-2 -n galera -- mysql -uroot -e <
查看数据:
# kubectl run mysql-client --image=mysql:5.7 -i -t --rm --restart=Never -- mysql -h 10.2.58.7 -e "SELECT * FROM demo.messages"If you don't see a command prompt, try pressing enter.+---------+| message |+---------+| hello |+---------+pod "mysql-client" deleted
定义集群内部访问数据库
如果pod之间互相访问,查询数据库就需要定义一个svc, 这里定义一个连接mysql的svc:
apiVersion: v1kind: Servicemetadata: name: mysql-read namespace: galera labels: app: mysqlspec: ports: - name: mysql port: 3306 selector: app: mysql
通过使用Pod来访问数据库:
# kubectl run mysql-client --image=mysql:5.7 -i -t --rm --restart=Never -- mysql -h mysql-read.galera -e "SELECT * FROM demo.messages"+---------+| message |+---------+| hello |+---------+pod "mysql-client" deleted
示例二: 部署mysql主从集群
官方参考文档
1. ceph集群中创建pool
在ceph 集群中创建一个kube的pool,用于数据库的存储池:
[root@ceph-1 ~]# ceph osd pool create kube 128pool 'kube' created
2. 使用之前创建的secretkey创建Storageclass
新定义一个storageclass:
apiVersion: storage.k8s.io/v1kind: StorageClassmetadata: name: mysqlprovisioner: kubernetes.io/rbdparameters: monitors: 192.168.20.41:6789,192.168.20.42:6789,192.168.20.43:6789 adminId: admin adminSecretName: ceph-secret pool: kube userId: admin userSecretName: ceph-secret fsType: xfs imageFormat: "2" imageFeatures: "layering"
3. 创建headless Service
由于要使用statefulSet进行主从数据库的部署,这里需要创建一个headless的service,和一个用于读库的service:
# Headless service for stable DNS entries of StatefulSet members.apiVersion: v1kind: Servicemetadata: name: mysql labels: app: mysqlspec: ports: - name: mysql port: 3306 clusterIP: None selector: app: mysql---# Client service for connecting to any MySQL instance for reads.# For writes, you must instead connect to the master: mysql-0.mysql.apiVersion: v1kind: Servicemetadata: name: mysql-read labels: app: mysqlspec: ports: - name: mysql port: 3306 selector: app: mysql
4. 创建用于主从同步的配置文件configmap
由于要进行主从同步,所以必须主库和从库必须要有相应的配置:
apiVersion: v1kind: ConfigMapmetadata: name: mysql labels: app: mysqldata: master.cnf: | # Apply this config only on the master. [mysqld] log-bin slave.cnf: | # Apply this config only on slaves. [mysqld] super-read-only
5 创建statefulSet
这里指定了使用StorageClass,使用RBD存储,同时需要使用一个xtrabackup的镜像进行数据同步:
apiVersion: apps/v1kind: StatefulSetmetadata: name: mysqlspec: selector: matchLabels: app: mysql serviceName: mysql replicas: 3 template: metadata: labels: app: mysql spec: initContainers: - name: init-mysql image: mysql:5.7 command: - bash - "-c" - | set -ex # Generate mysql server-id from pod ordinal index. [[ `hostname` =~ -([0-9]+)$ ]] || exit 1 ordinal=${BASH_REMATCH[1]} echo [mysqld] > /mnt/conf.d/server-id.cnf # Add an offset to avoid reserved server-id=0 value. echo server-id=$((100 + $ordinal)) >> /mnt/conf.d/server-id.cnf # Copy appropriate conf.d files from config-map to emptyDir. if [[ $ordinal -eq 0 ]]; then cp /mnt/config-map/master.cnf /mnt/conf.d/ else cp /mnt/config-map/slave.cnf /mnt/conf.d/ fi volumeMounts: - name: conf mountPath: /mnt/conf.d - name: config-map mountPath: /mnt/config-map - name: clone-mysql image: tangup/xtrabackup:1.0 command: - bash - "-c" - | set -ex # Skip the clone if data already exists. [[ -d /var/lib/mysql/mysql ]] && exit 0 # Skip the clone on master (ordinal index 0). [[ `hostname` =~ -([0-9]+)$ ]] || exit 1 ordinal=${BASH_REMATCH[1]} [[ $ordinal -eq 0 ]] && exit 0 # Clone data from previous peer. ncat --recv-only mysql-$(($ordinal-1)).mysql 3307 | xbstream -x -C /var/lib/mysql # Prepare the backup. xtrabackup --prepare --target-dir=/var/lib/mysql volumeMounts: - name: data mountPath: /var/lib/mysql subPath: mysql - name: conf mountPath: /etc/mysql/conf.d containers: - name: mysql image: mysql:5.7 env: - name: MYSQL_ALLOW_EMPTY_PASSWORD value: "1" ports: - name: mysql containerPort: 3306 volumeMounts: - name: data mountPath: /var/lib/mysql subPath: mysql - name: conf mountPath: /etc/mysql/conf.d resources: requests: cpu: 500m memory: 1Gi livenessProbe: exec: command: ["mysqladmin", "ping"] initialDelaySeconds: 30 periodSeconds: 10 timeoutSeconds: 5 readinessProbe: exec: # Check we can execute queries over TCP (skip-networking is off). command: ["mysql", "-h", "127.0.0.1", "-e", "SELECT 1"] initialDelaySeconds: 5 periodSeconds: 2 timeoutSeconds: 1 - name: xtrabackup image: tangup/xtrabackup:1.0 ports: - name: xtrabackup containerPort: 3307 command: - bash - "-c" - | set -ex cd /var/lib/mysql # Determine binlog position of cloned data, if any. if [[ -f xtrabackup_slave_info ]]; then # XtraBackup already generated a partial "CHANGE MASTER TO" query # because we're cloning from an existing slave. mv xtrabackup_slave_info change_master_to.sql.in # Ignore xtrabackup_binlog_info in this case (it's useless). rm -f xtrabackup_binlog_info elif [[ -f xtrabackup_binlog_info ]]; then # We're cloning directly from master. Parse binlog position. [[ `cat xtrabackup_binlog_info` =~ ^(.*?)[[:space:]]+(.*?)$ ]] || exit 1 rm xtrabackup_binlog_info echo "CHANGE MASTER TO MASTER_LOG_FILE='${BASH_REMATCH[1]}',\ MASTER_LOG_POS=${BASH_REMATCH[2]}" > change_master_to.sql.in fi # Check if we need to complete a clone by starting replication. if [[ -f change_master_to.sql.in ]]; then echo "Waiting for mysqld to be ready (accepting connections)" until mysql -h 127.0.0.1 -e "SELECT 1"; do sleep 1; done echo "Initializing replication from clone position" # In case of container restart, attempt this at-most-once. mv change_master_to.sql.in change_master_to.sql.orig mysql -h 127.0.0.1 <
6. 检查集群状态
查看pod:
[root@master-1 ~]# kubectl get poNAME READY STATUS RESTARTS AGEmysql-0 2/2 Running 2 110mmysql-1 2/2 Running 0 109mmysql-2 2/2 Running 0 16m
pvc:
[root@master-1 ~]# kubectl get pvc |grep mysql|grep -v fastdata-mysql-0 Bound pvc-3737108a-6a2a-11e9-ac56-000c296b46ac 1Gi RWO mysql 5h53mdata-mysql-1 Bound pvc-279bdca0-6a4a-11e9-ac56-000c296b46ac 1Gi RWO mysql 114mdata-mysql-2 Bound pvc-fbe153bc-6a52-11e9-ac56-000c296b46ac 1Gi RWO mysql 51m
Ceph集群上自动创建的镜像:
[root@ceph-1 ~]# rbd list kubekubernetes-dynamic-pvc-2ee47370-6a4a-11e9-bb82-000c296b46ackubernetes-dynamic-pvc-39a42869-6a2a-11e9-bb82-000c296b46ackubernetes-dynamic-pvc-fbead120-6a52-11e9-bb82-000c296b46ac
7.测试数据库集群
向主库写入数据,使用headless server所提供的 podname.headlessname 的形式就可以直接访问POD, 这在DNS解析中是固定的。这里访问mysql-0就使用mysql-0.mysql:
kubectl run mysql-client --image=mysql:5.7 -i --rm --restart=Never --\ mysql -h mysql-0.mysql <
使用mysql-read去访问数据库数据:
# kubectl run mysql-client --image=mysql:5.7 -i -t --rm --restart=Never -- mysql -h mysql-read -e "SELECT * FROM test.messages"+---------+| message |+---------+| hello |+---------+
可以使用如下命令去循环的查看当前是mysql-read连接的数据库:
kubectl run mysql-client-loop --image=mysql:5.7 -i -t --rm --restart=Never --\ bash -ic "while sleep 1; do mysql -h mysql-read -e 'SELECT @@server_id,NOW()'; done" +-------------+---------------------+| @@server_id | NOW() |+-------------+---------------------+| 102 | 2019-04-28 20:24:11 |+-------------+---------------------++-------------+---------------------+| @@server_id | NOW() |+-------------+---------------------+| 101 | 2019-04-28 20:27:35 |+-------------+---------------------++-------------+---------------------+| @@server_id | NOW() |+-------------+---------------------+| 100 | 2019-04-28 20:18:38 |+-------------+---------------------+