1. Standalone：在 K8S 启动一个长期运行的集群，所有 Job 都通过 spark-submit 向这个集群提交
2. Kubernetes Native：通过 spark-submit 直接向 K8S 的 API Server 提交，申请到资源后启动 Pod 做为 Driver 和 Executor 执行 Job，参考 http://spark.apache.org/docs/2.4.6/running-on-kubernetes.html
3.Spark Operator：安装 Spark Operator，然后定义 spark-app.yaml，再执行 kubectl apply -f spark-app.yaml，这种申明式 API 和调用方式是 K8S 的典型应用方式，参考 https://github.com/GoogleCloudPlatform/spark-on-k8s-operator

下载spark包

wget https://dlcdn.apache.org/spark/spark-3.2.0/spark-3.2.0-bin-hadoop2.7.tgz
tar -zxvf 	spark-3.2.0-bin-hadoop2.7.tgz

构建镜像:

共三个镜像， java/scala共一个、  python_image、  R_image
./bin/docker-image-tool.sh -t 3.2 build
./bin/docker-image-tool.sh -t 3.2 -p kubernetes/dockerfiles/spark/bindings/python/Dockerfile build

提交任务

使用root@anneng-730xd

cd /ceph_disk1/software/spark-3.2.0-bin-hadoop2.7

./bin/spark-submit \
    --master k8s://https://192.168.1.2:6443 \
    --deploy-mode cluster \
    --name spark-pi \
    --class org.apache.spark.examples.SparkPi \
    --conf spark.executor.instances=5 \
    --conf spark.kubernetes.container.image=spark:3.2 \
    --conf spark.kubernetes.authenticate.driver.serviceAccountName=spark \
    local:///opt/spark/examples/jars/spark-examples_2.12-3.1.1.jar
    
此处local:// 后的地址为容器内路径


bin/spark-submit \
    --master k8s://https://192.168.1.2:6443 \
    --deploy-mode cluster \
    --name spark-test \
    --conf spark.executor.instances=3 \
    --conf spark.kubernetes.container.image=spark-py:3.2 \
    --conf spark.kubernetes.authenticate.driver.serviceAccountName=spark \
    local:///opt/spark/examples/src/main/python/wordcount.py \
    /opt/spark/examples/src/main/python/wordcount.py