K8S应用笔记 —— 部署Dolphinscheduler及简单应用（一）-Toy模板网

这篇具有很好参考价值的文章主要介绍了K8S应用笔记 —— 部署Dolphinscheduler及简单应用（一）。希望对大家有所帮助。如果存在错误或未考虑完全的地方，请大家不吝赐教，您也可以点击"举报违法"按钮提交疑问。

一、Dolphinscheduler简介

Apache DolphinScheduler 是一个分布式易扩展的可视化DAG工作流任务调度开源系统。适用于企业级场景，提供了一个可视化操作任务、工作流和全生命周期数据处理过程的解决方案。

Apache DolphinScheduler 旨在解决复杂的大数据任务依赖关系，并为应用程序提供数据和各种 OPS 编排中的关系。解决数据研发ETL依赖错综复杂，无法监控任务健康状态的问题。 DolphinScheduler 以 DAG（Directed Acyclic Graph，DAG）流式方式组装任务，可以及时监控任务的执行状态，支持重试、指定节点恢复失败、暂停、恢复、终止任务等操作。

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

二、本章目标

基于K8S环境完成Dolphinscheduler部署
使用本地文件存储而非 HDFS 和 S3
基于K8S环境Dolphinscheduler简单应用（支持Python3和MySQL数据源及工作流编排）

三、前提条件

具备 Kubernetes 1.12+ 集群（必须），使用Kuboard v3作为集群管理工具（可选）

具体操作可见：
K8S安装笔记（一）—— master节点完整安装配置
K8S安装笔记（二）—— 多公网服务器搭建集群
PV 供应（存储使用NFS，存储类为nfs-storage）

具体操作可见：
搭建NFS Server及创建 NFS 存储类

四、安装helm

helm官方文档，https://helm.sh/docs/intro/install/

4.1 下载所需版本

下载路径：https://github.com/helm/helm/releases

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

我选择的选择的版本是：helm-v3.12.2-linux-amd64

4.2 上传至服务器并解压

tar -zxvf helm-v3.12.2-linux-amd64.tar.gz

4.3 移到到可执行目录

helm在解压后的目录中找到二进制文件，然后将其移至所需的目标位置

mv linux-amd64/helm /usr/local/bin/helm

4.4 从脚本安装

安装脚本：

curl -fsSL -o get_helm.sh https://raw.githubusercontent.com/helm/helm/master/scripts/get-helm-3
chmod 700 get_helm.sh
./get_helm.sh

五、安装 dolphinscheduler

5.1 下载解压

请下载源码包 apache-dolphinscheduler--src.tar.gz，下载地址。

发布一个名为 dolphinscheduler 的版本(release)，官方给出的参照如下：

tar -zxvf apache-dolphinscheduler-<version>-src.tar.gz
cd apache-dolphinscheduler-<version>-src/deploy/kubernetes/dolphinscheduler
helm repo add bitnami https://charts.bitnami.com/bitnami
helm dependency update .
helm install dolphinscheduler . --set image.tag=<version>

我选择的是3.1.8版本，具体执行命令如下：

tar -zxvf apache-dolphinscheduler-3.1.8-src.tar.gz
cd apache-dolphinscheduler-3.1.8-src/deploy/kubernetes/dolphinscheduler

5.2 变更RAW资源请求地址

原请求地址为：https://raw.githubusercontent.com/，国内可能连接不上，变更RAW资源加速地址：https://raw.gitmirror.com

修改Chart.yaml文件中下面的配置项：

vi Chart.yaml
#原地址
#repository: https://raw.githubusercontent.com/bitnami/charts/archive-full-index/bitnami
#资源加速地址
repository: https://raw.gitmirror.com/bitnami/charts/archive-full-index/bitnami

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

5.3 支持本地文件存储而非 HDFS和S3

修改 values.yaml 文件中下面的配置项：

common:
  configmap:
    RESOURCE_STORAGE_TYPE: "NONE"
    RESOURCE_UPLOAD_PATH: "/dolphinscheduler"
    FS_DEFAULT_FS: "file:///"
  fsFileResourcePersistence:
    enabled: true
    accessModes:
    - "ReadWriteMany"
    storageClassName: "-"
    storage: "20Gi"

storageClassName 和 storage 按需修改为实际值，注意：storageClassName 必须支持访问模式：ReadWriteMany。

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

5.4 部署

helm dependency update .
helm install dolphinscheduler . -n dolphinscheduler

部署效果：

kubectl get pod,svc,pvc -o wide -n dolphinscheduler

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

5.5 访问前端页面

port-forward 端口转发：

kubectl port-forward --address 0.0.0.0 -n dolphinscheduler svc/dolphinscheduler-api 12345:12345

其他NodePort或ingress访问方式请自行探索。

访问前端页面：http://localhost:12345/dolphinscheduler/ui

默认的用户是admin，默认的密码是dolphinscheduler123

六、支持Python3和MySQL

6.1 dolphinscheduler-worker 镜像构建

下载 MySQL 驱动包 mysql-connector-java-8.0.16.jar

目录结构：
dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

创建一个新的 Dockerfile，用于添加 MySQL 的驱动包和安装Python 3：

FROM dolphinscheduler.docker.scarf.sh/apache/dolphinscheduler-worker:3.1.8
# 添加mysql驱动
COPY ./mysql-connector-java-8.0.16.jar /opt/dolphinscheduler/libs
# 添加自定义requirements.txt
COPY ./requirements.txt /tmp
RUN apt-get update && \
    apt-get install -y --no-install-recommends python3-pip && \
	pip3 install --no-cache-dir -r /tmp/requirements.txt -i https://mirrors.aliyun.com/pypi/simple/ && \
    rm -rf /var/lib/apt/lists/*

requirements.txt：

Flask_Cors==3.0.10
pandas==1.4.2
PyMySQL==1.0.2
SQLAlchemy==1.4.32
xlwt==1.3.0
xlsxwriter==3.0.3
gunicorn
greenlet
eventlet
gevent
pypinyin
openpyxl

构建一个包含新镜像：

docker build -t apache/dolphinscheduler-worker:python3-mysql .

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

将构建好的新镜像进行分发。

6.2 dolphinscheduler-api 镜像构建

创建一个新的 Dockerfile，用于添加 MySQL 的驱动包和安装Python 3：

FROM dolphinscheduler.docker.scarf.sh/apache/dolphinscheduler-api:3.1.8
# 添加mysql驱动

构建一个包含新镜像：

docker build -t apache/dolphinscheduler-api:support-mysql .

将构建好的新镜像进行分发。

6.3 修改PYTHON_HOME

修改 values.yaml 文件中的 PYTHON_HOME 为 /usr/bin/python3
或在Kuboard - dolphinscheduler 名称空间 - 配置中心 - 配置字典 - dolphinscheduler-common - 编辑

6.4 修改dolphinscheduler-worker 运行镜像版本

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

6.5 修改dolphinscheduler-api 运行镜像版本

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

七、简单使用验证

7.1 登录

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

7.2 创建租户

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

7.3 创建项目

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

7.4 创建数据源

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

7.5 创建项目

dolphinscheduler k8s,Kubernetes实战,kubernetes,容器,云原生

7.6 创建工作流

[{
	"processDefinition": {
		"id": 1,
		"code": 10577001612288,
		"name": "测试",
		"version": 1,
		"releaseState": "OFFLINE",
		"projectCode": 10576969989760,
		"description": "",
		"globalParams": "[]",
		"globalParamList": [],
		"globalParamMap": {},
		"createTime": "2023-08-15 09:33:45",
		"updateTime": "2023-08-15 09:33:45",
		"flag": "YES",
		"userId": 1,
		"userName": null,
		"projectName": null,
		"locations": "[{\"taskCode\":10576976496512,\"x\":175,\"y\":216},{\"taskCode\":10576980784256,\"x\":498,\"y\":216},{\"taskCode\":10576997351040,\"x\":834,\"y\":216}]",
		"scheduleReleaseState": null,
		"timeout": 0,
		"tenantId": -1,
		"tenantCode": null,
		"modifyBy": null,
		"warningGroupId": 0,
		"executionType": "PARALLEL"
	},
	"processTaskRelationList": [{
		"id": 1,
		"name": "",
		"processDefinitionVersion": 1,
		"projectCode": 10576969989760,
		"processDefinitionCode": 10577001612288,
		"preTaskCode": 0,
		"preTaskVersion": 0,
		"postTaskCode": 10576976496512,
		"postTaskVersion": 1,
		"conditionType": "NONE",
		"conditionParams": {},
		"createTime": "2023-08-15 09:33:45",
		"updateTime": "2023-08-15 09:33:45",
		"operator": 1,
		"operateTime": "2023-08-15 09:33:45"
	}, {
		"id": 2,
		"name": "",
		"processDefinitionVersion": 1,
		"projectCode": 10576969989760,
		"processDefinitionCode": 10577001612288,
		"preTaskCode": 10576976496512,
		"preTaskVersion": 1,
		"postTaskCode": 10576980784256,
		"postTaskVersion": 1,
		"conditionType": "NONE",
		"conditionParams": {},
		"createTime": "2023-08-15 09:33:45",
		"updateTime": "2023-08-15 09:33:45",
		"operator": 1,
		"operateTime": "2023-08-15 09:33:45"
	}, {
		"id": 3,
		"name": "",
		"processDefinitionVersion": 1,
		"projectCode": 10576969989760,
		"processDefinitionCode": 10577001612288,
		"preTaskCode": 10576980784256,
		"preTaskVersion": 1,
		"postTaskCode": 10576997351040,
		"postTaskVersion": 1,
		"conditionType": "NONE",
		"conditionParams": {},
		"createTime": "2023-08-15 09:33:45",
		"updateTime": "2023-08-15 09:33:45",
		"operator": 1,
		"operateTime": "2023-08-15 09:33:45"
	}],
	"taskDefinitionList": [{
		"id": 1,
		"code": 10576976496512,
		"name": "sql",
		"version": 1,
		"description": "",
		"projectCode": 10576969989760,
		"userId": 1,
		"taskType": "SQL",
		"taskParams": {
			"localParams": [],
			"resourceList": [],
			"type": "MYSQL",
			"datasource": 1,
			"sql": "show tables",
			"sqlType": "0",
			"preStatements": [],
			"postStatements": [],
			"segmentSeparator": "",
			"displayRows": 10
		},
		"taskParamList": [],
		"taskParamMap": null,
		"flag": "YES",
		"taskPriority": "MEDIUM",
		"userName": null,
		"projectName": null,
		"workerGroup": "default",
		"environmentCode": -1,
		"failRetryTimes": 0,
		"failRetryInterval": 1,
		"timeoutFlag": "CLOSE",
		"timeoutNotifyStrategy": null,
		"timeout": 0,
		"delayTime": 0,
		"resourceIds": "",
		"createTime": "2023-08-15 09:33:45",
		"updateTime": "2023-08-15 09:33:45",
		"modifyBy": null,
		"taskGroupId": 0,
		"taskGroupPriority": 0,
		"cpuQuota": -1,
		"memoryMax": -1,
		"taskExecuteType": "BATCH",
		"operator": 1,
		"operateTime": "2023-08-15 09:33:45"
	}, {
		"id": 2,
		"code": 10576980784256,
		"name": "shell",
		"version": 1,
		"description": "",
		"projectCode": 10576969989760,
		"userId": 1,
		"taskType": "SHELL",
		"taskParams": {
			"localParams": [],
			"rawScript": "echo \"hello shell\"",
			"resourceList": []
		},
		"taskParamList": [],
		"taskParamMap": null,
		"flag": "YES",
		"taskPriority": "MEDIUM",
		"userName": null,
		"projectName": null,
		"workerGroup": "default",
		"environmentCode": -1,
		"failRetryTimes": 0,
		"failRetryInterval": 1,
		"timeoutFlag": "CLOSE",
		"timeoutNotifyStrategy": null,
		"timeout": 0,
		"delayTime": 0,
		"resourceIds": "",
		"createTime": "2023-08-15 09:33:45",
		"updateTime": "2023-08-15 09:33:45",
		"modifyBy": null,
		"taskGroupId": 0,
		"taskGroupPriority": 0,
		"cpuQuota": -1,
		"memoryMax": -1,
		"taskExecuteType": "BATCH",
		"operator": 1,
		"operateTime": "2023-08-15 09:33:45"
	}, {
		"id": 3,
		"code": 10576997351040,
		"name": "python",
		"version": 1,
		"description": "",
		"projectCode": 10576969989760,
		"userId": 1,
		"taskType": "PYTHON",
		"taskParams": {
			"localParams": [],
			"rawScript": "print(\"hello python\")",
			"resourceList": []
		},
		"taskParamList": [],
		"taskParamMap": null,
		"flag": "YES",
		"taskPriority": "MEDIUM",
		"userName": null,
		"projectName": null,
		"workerGroup": "default",
		"environmentCode": -1,
		"failRetryTimes": 0,
		"failRetryInterval": 1,
		"timeoutFlag": "CLOSE",
		"timeoutNotifyStrategy": null,
		"timeout": 0,
		"delayTime": 0,
		"resourceIds": "",
		"createTime": "2023-08-15 09:33:45",
		"updateTime": "2023-08-15 09:33:45",
		"modifyBy": null,
		"taskGroupId": 0,
		"taskGroupPriority": 0,
		"cpuQuota": -1,
		"memoryMax": -1,
		"taskExecuteType": "BATCH",
		"operator": 1,
		"operateTime": "2023-08-15 09:33:45"
	}],
	"schedule": null
}]