参考：【云原生之kubernetes实战】在k8s环境下部署Spark分布式计算平台 - 小鸟云

 

一、Spark介绍

1.Spark简介

Spark是分布式计算平台，是一个用scala语言编写的计算框架，基于内存的快速、通用、可扩展的大数据分析引擎。

2.Spark作用

Apache Spark 是一个快速的，通用的集群计算系统。它对 Java，Scala，Python 和 R 提供了的高层 API，并有一个经优化的支持通用执行图计算的引擎。它还支持一组丰富的高级工具，包括用于 SQL 和结构化数据处理的 Spark SQL，用于机器学习的 MLlib，用于图计算的 GraphX 和 Spark Streaming。

二、检查本地集群状态

1.检查工作节点状态

[root@master ~]# kubectl get nodes 
NAME     STATUS   ROLES           AGE   VERSION
master   Ready    control-plane   19d   v1.24.0
node01   Ready    <none>          19d   v1.24.0
node02   Ready    <none>          19d   v1.24.0

2.检查k8s版本

[root@master ~]# kubectl version --short
Flag --short has been deprecated, and will be removed in the future. The --short output will become the default.
Client Version: v1.24.0
Kustomize Version: v4.5.4
Server Version: v1.24.0

二、安装helm工具

1.下载helm软件包

[root@master mysql]# wget https://get.helm.sh/helm-v3.9.0-linux-amd64.tar.gz
--2022-10-22 19:10:12--  https://get.helm.sh/helm-v3.9.0-linux-amd64.tar.gz
Resolving get.helm.sh (get.helm.sh)... 152.199.39.108, 2606:2800:247:1cb7:261b:1f9c:2074:3c
Connecting to get.helm.sh (get.helm.sh)|152.199.39.108|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 13952532 (13M) [application/x-tar]
Saving to: ‘helm-v3.9.0-linux-amd64.tar.gz’

100%[========================================================================================>] 13,952,532  16.7MB/s   in 0.8s   

2022-10-22 19:10:17 (16.7 MB/s) - ‘helm-v3.9.0-linux-amd64.tar.gz’ saved [13952532/13952532]

2.解压压缩包

[root@master mysql]# tar -xzf helm-v3.9.0-linux-amd64.tar.gz
[root@master mysql]# ls
helm-v3.9.0-linux-amd64.tar.gz  linux-amd64

3.复制二进制文件

[root@master linux-amd64]# ls
helm  LICENSE  README.md
[root@master linux-amd64]# cp -a helm /usr/bin/
[root@master linux-amd64]# 

4.检查helm版本

[root@master linux-amd64]# helm version
version.BuildInfo{Version:"v3.9.0", GitCommit:"7ceeda6c585217a19a1131663d8cd1f7d641b2a7", GitTreeState:"clean", GoVersion:"go1.17.5"}

5.helm命令补全

[root@master spark]# helm completion bash > .helmrc && echo "source .helmrc" >> .bashrc
[root@master mysql]# source .helmrc
[root@master mysql]# 

三、安装nfs服务器

1.安装nfs软件

 yum install -y nfs-utils

2.创建共享目录

mkdir -p /nfs && chmod 766 -R /nfs

3配置共享目录

echo "/nfs/ *(insecure,rw,sync,no_root_squash)" > /etc/exports

4.使nfs配置生效

exportfs -r

5.设置nfs服务开机自启

 systemctl enable --now rpcbind
 systemctl enable --now  nfs-server

6.其他节点检查nfs共享情况

[root@node01 ~]# showmount -e 192.168.3.90
Export list for 192.168.3.90:
/nfs *

四、部署storageclass

1.编辑sc.yaml

[root@master spark]# cat sc.yaml 
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
 name: nfs-storage
 annotations:
   storageclass.kubernetes.io/is-default-class: "true"
provisioner: k8s-sigs.io/nfs-subdir-external-provisioner
parameters:
 archiveOnDelete: "true"  ## 删除pv的时候，pv的内容是否要备份

---
apiVersion: apps/v1
kind: Deployment
metadata:
 name: nfs-client-provisioner
 labels:
   app: nfs-client-provisioner
 # replace with namespace where provisioner is deployed
 namespace: default
spec:
 replicas: 1
 strategy:
   type: Recreate
 selector:
   matchLabels:
     app: nfs-client-provisioner
 template:
   metadata:
     labels:
       app: nfs-client-provisioner
   spec:
     serviceAccountName: nfs-client-provisioner
     containers:
       - name: nfs-client-provisioner
         image: registry.cn-hangzhou.aliyuncs.com/lfy_k8s_images/nfs-subdir-external-provisioner:v4.0.2
         # resources:
         #    limits:
         #      cpu: 10m
         #    requests:
         #      cpu: 10m
         volumeMounts:
           - name: nfs-client-root
             mountPath: /persistentvolumes
         env:
           - name: PROVISIONER_NAME
             value: k8s-sigs.io/nfs-subdir-external-provisioner
           - name: NFS_SERVER
             value: 192.168.3.90 ## 指定自己nfs服务器地址
           - name: NFS_PATH  
             value: /nfs  ## nfs服务器共享的目录
     volumes:
       - name: nfs-client-root
         nfs:
           server: 192.168.3.90
           path: /nfs
---
apiVersion: v1
kind: ServiceAccount
metadata:
 name: nfs-client-provisioner
 # replace with namespace where provisioner is deployed
 namespace: default
---
kind: ClusterRole
apiVersion: rbac.authorization.k8s.io/v1
metadata:
 name: nfs-client-provisioner-runner
rules:
 - apiGroups: [""]
   resources: ["nodes"]
   verbs: ["get", "list", "watch"]
 - apiGroups: [""]
   resources: ["persistentvolumes"]
   verbs: ["get", "list", "watch", "create", "delete"]
 - apiGroups: [""]
   resources: ["persistentvolumeclaims"]
   verbs: ["get", "list", "watch", "update"]
 - apiGroups: ["storage.k8s.io"]
   resources: ["storageclasses"]
   verbs: ["get", "list", "watch"]
 - apiGroups: [""]
   resources: ["events"]
   verbs: ["create", "update", "patch"]
---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
 name: run-nfs-client-provisioner
subjects:
 - kind: ServiceAccount
   name: nfs-client-provisioner
   # replace with namespace where provisioner is deployed
   namespace: default
roleRef:
 kind: ClusterRole
 name: nfs-client-provisioner-runner
 apiGroup: rbac.authorization.k8s.io
---
kind: Role
apiVersion: rbac.authorization.k8s.io/v1
metadata:
 name: leader-locking-nfs-client-provisioner
 # replace with namespace where provisioner is deployed
 namespace: default
rules:
 - apiGroups: [""]
   resources: ["endpoints"]
   verbs: ["get", "list", "watch", "create", "update", "patch"]
---
kind: RoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
 name: leader-locking-nfs-client-provisioner
 # replace with namespace where provisioner is deployed
 namespace: default
subjects:
 - kind: ServiceAccount
   name: nfs-client-provisioner
   # replace with namespace where provisioner is deployed
   namespace: default
roleRef:
 kind: Role
 name: leader-locking-nfs-client-provisioner
 apiGroup: rbac.authorization.k8s.io

2.应用sc.yaml文件

[root@master spark]# kubectl apply -f sc.yaml 
storageclass.storage.k8s.io/nfs-storage created
deployment.apps/nfs-client-provisioner created
serviceaccount/nfs-client-provisioner created
clusterrole.rbac.authorization.k8s.io/nfs-client-provisioner-runner created
clusterrolebinding.rbac.authorization.k8s.io/run-nfs-client-provisioner created
role.rbac.authorization.k8s.io/leader-locking-nfs-client-provisioner created
rolebinding.rbac.authorization.k8s.io/leader-locking-nfs-client-provisioner created

3.检查storageclass资源对象状态

[root@master spark]# kubectl get storageclasses.storage.k8s.io 
NAME                    PROVISIONER                                   RECLAIMPOLICY   VOLUMEBINDINGMODE   ALLOWVOLUMEEXPANSION   AGE
nfs-storage (default)   k8s-sigs.io/nfs-subdir-external-provisioner   Delete          Immediate           false                  81s

五、添加helm仓库源

1.添加helm仓库

[root@master spark]# helm repo add bitnami https://charts.bitnami.com/bitnami/
"bitnami" has been added to your repositories

2.查看helm仓库列表

[root@master spark]# helm repo list 
NAME     	URL                                                                      
bitnami  	https://charts.bitnami.com/bitnami                                       
azure    	http://mirror.azure.cn/kubernetes/charts/                                
incubator	https://aliacs-app-catalog.oss-cn-hangzhou.aliyuncs.com/charts-incubator/

3.更新helm仓库

[root@master spark]#  helm repo update
Hang tight while we grab the latest from your chart repositories...
...Successfully got an update from the "incubator" chart repository
...Successfully got an update from the "azure" chart repository
...Successfully got an update from the "bitnami" chart repository
Update Complete. ⎈Happy Helming!⎈

4.搜索关于Spark的资源

[root@master spark]# helm search repo spark
NAME                              	CHART VERSION	APP VERSION	DESCRIPTION                                       
azure/spark                       	1.0.5        	1.5.1      	DEPRECATED - Fast and general-purpose cluster c...
azure/spark-history-server        	1.4.3        	2.4.0      	DEPRECATED - A Helm chart for Spark History Server
bitnami/spark                     	6.3.6        	3.3.0      	Apache Spark is a high-performance engine for l...
incubator/ack-spark-history-server	0.5.0        	2.4.5      	A Helm chart for Spark History Server             
incubator/ack-spark-operator      	0.1.16       	2.4.5      	A Helm chart for Spark on Kubernetes operator     
bitnami/dataplatform-bp1          	12.0.2       	1.0.1      	DEPRECATED This Helm chart can be used for the ...
bitnami/dataplatform-bp2          	12.0.5       	1.0.1      	DEPRECATED This Helm chart can be used for the ...
azure/luigi                       	2.7.8        	2.7.2      	DEPRECATED Luigi is a Python module that helps ...

六、安装Spark

1.下载chart

[root@master spark]# helm pull bitnami/spark
[root@master spark]# ls
spark-6.3.6.tgz
[root@master spark]# tar -xzf spark-6.3.6.tgz 
[root@master spark]# ls
spark  spark-6.3.6.tgz

2.修改values.yaml

修改部分

service:
  ## @param service.type Kubernetes Service type
  ##
  type: NodePort
  ## @param service.ports.http Spark client port for HTTP
  ## @param service.ports.https Spark client port for HTTPS
  ## @param service.ports.cluster Spark cluster port
  ##
  ports:
    http: 80
    https: 443
    cluster: 7077
  ## Specify the nodePort(s) value(s) for the LoadBalancer and NodePort service types.
  ## ref: https://kubernetes.io/docs/concepts/services-networking/service/#type-nodeport
  ## @param service.nodePorts.http Kubernetes web node port for HTTP
  ## @param service.nodePorts.https Kubernetes web node port for HTTPS
  ## @param service.nodePorts.cluster Kubernetes cluster node port
  ##
  nodePorts:
              

3.helm安装Spark应用

[root@master spark]# helm install myspark ./spark
NAME: myspark
LAST DEPLOYED: Sun Oct 23 00:05:40 2022
NAMESPACE: default
STATUS: deployed
REVISION: 1
TEST SUITE: None
NOTES:
CHART NAME: spark
CHART VERSION: 6.3.6
APP VERSION: 3.3.0

** Please be patient while the chart is being deployed **

1. Get the Spark master WebUI URL by running these commands:

  export NODE_PORT=$(kubectl get --namespace default -o jsonpath="{.spec.ports[?(@.name=='http')].nodePort}" services myspark-master-svc)
  export NODE_IP=$(kubectl get nodes --namespace default -o jsonpath="{.items[0].status.addresses[0].address}")
  echo http://$NODE_IP:$NODE_PORT

2. Submit an application to the cluster:

  To submit an application to the cluster the spark-submit script must be used. That script can be
  obtained at https://github.com/apache/spark/tree/master/bin. Also you can use kubectl run.

  Run the commands below to obtain the master IP and submit your application.

  export EXAMPLE_JAR=$(kubectl exec -ti --namespace default myspark-worker-0 -- find examples/jars/ -name 'spark-example*\.jar' | tr -d '\r')
  export SUBMIT_PORT=$(kubectl get --namespace default -o jsonpath="{.spec.ports[?(@.name=='cluster')].nodePort}" services myspark-master-svc)
  export SUBMIT_IP=$(kubectl get nodes --namespace default -o jsonpath="{.items[0].status.addresses[0].address}")

  kubectl run --namespace default myspark-client --rm --tty -i --restart='Never' \
    --image docker.io/bitnami/spark:3.3.0-debian-11-r40 \
    -- spark-submit --master spark://$SUBMIT_IP:$SUBMIT_PORT \
    --class org.apache.spark.examples.SparkPi \
    --deploy-mode cluster \
    $EXAMPLE_JAR 1000

2.检查pod状态

[root@master spark]# kubectl get pod
NAME                                     READY   STATUS    RESTARTS        AGE
my-tomcat9                               1/1     Running   2 (5h55m ago)   19d
myspark-master-0                         1/1     Running   0               36m
myspark-worker-0                         1/1     Running   0               36m
myspark-worker-1                         1/1     Running   0               33m
nfs-client-provisioner-8dcd8c766-2bptf   1/1     Running   0               5h16m

3.检查svc

[root@master spark]# kubectl get svc
NAME                   TYPE        CLUSTER-IP    EXTERNAL-IP   PORT(S)                       AGE
kubernetes             ClusterIP   10.96.0.1     <none>        443/TCP                       20d
myspark-headless       ClusterIP   None          <none>        <none>                        36m
myspark-master-svc     NodePort    10.96.2.220   <none>        7077:32573/TCP,80:31295/TCP   36m

4.删除spark应用

helm delete --purge myspark

七、访问Spark的Web UI

worker 实例当前为 2 个。

八、新增worker 实例数量

1.修改values.yaml

更改values.yaml中 replicaCount为3

 replicaCount: 3
  ## Kubernetes Pods Security Context
  ## https://kubernetes.io/docs/tasks/configure-pod-container/security-context/
  ## @param worker.podSecurityContext.enabled Enable security context
  ## @param worker.podSecurityContext.fsGroup Group ID for the container
  ## @param worker.podSecurityContext.runAsUser User ID for the container
  ## @param worker.podSecurityContext.runAsGroup Group ID for the container
  ## @param worker.podSecurityContext.seLinuxOptions SELinux options for the container
  ##
  podSecurityContext:
    enabled: true
    fsGroup: 1001
    runAsUser: 1001
    runAsGroup: 0
    seLinuxOptions: {}

2.使用helm更新spark应用

[root@master spark]# helm upgrade myspark  ./spark
Release "myspark" has been upgraded. Happy Helming!
NAME: myspark
LAST DEPLOYED: Sun Oct 23 00:52:36 2022
NAMESPACE: default
STATUS: deployed
REVISION: 3
TEST SUITE: None
NOTES:
CHART NAME: spark
CHART VERSION: 6.3.6
APP VERSION: 3.3.0

** Please be patient while the chart is being deployed **

1. Get the Spark master WebUI URL by running these commands:

  export NODE_PORT=$(kubectl get --namespace default -o jsonpath="{.spec.ports[?(@.name=='http')].nodePort}" services myspark-master-svc)
  export NODE_IP=$(kubectl get nodes --namespace default -o jsonpath="{.items[0].status.addresses[0].address}")
  echo http://$NODE_IP:$NODE_PORT

2. Submit an application to the cluster:

  To submit an application to the cluster the spark-submit script must be used. That script can be
  obtained at https://github.com/apache/spark/tree/master/bin. Also you can use kubectl run.

  Run the commands below to obtain the master IP and submit your application.

  export EXAMPLE_JAR=$(kubectl exec -ti --namespace default myspark-worker-0 -- find examples/jars/ -name 'spark-example*\.jar' | tr -d '\r')
  export SUBMIT_PORT=$(kubectl get --namespace default -o jsonpath="{.spec.ports[?(@.name=='cluster')].nodePort}" services myspark-master-svc)
  export SUBMIT_IP=$(kubectl get nodes --namespace default -o jsonpath="{.items[0].status.addresses[0].address}")

  kubectl run --namespace default myspark-client --rm --tty -i --restart='Never' \
    --image docker.io/bitnami/spark:3.3.0-debian-11-r40 \
    -- spark-submit --master spark://$SUBMIT_IP:$SUBMIT_PORT \
    --class org.apache.spark.examples.SparkPi \
    --deploy-mode cluster \
    $EXAMPLE_JAR 1000

3.检查pod状态

[root@master spark]# kubectl get pods
NAME                                     READY   STATUS    RESTARTS       AGE
my-tomcat9                               1/1     Running   2 (6h7m ago)   20d
my-wordpress-9585b7f4d-5lfzn             1/1     Running   1 (78m ago)    82m
my-wordpress-mariadb-0                   1/1     Running   0              82m
myspark-master-0                         1/1     Running   0              48m
myspark-worker-0                         1/1     Running   0              48m
myspark-worker-1                         1/1     Running   0              45m
myspark-worker-2                         1/1     Running   0              82s
nfs-client-provisioner-8dcd8c766-2bptf   1/1     Running   0              5h28m

4.查看spark的Web UI中worker数量