dolphinscheduler调度器接入注意事项等信息可参考我的上一篇博客进行了解，地址在这里 ->
@[TOC]

一、功能清单

二、执行/停止任务

说明：
大数据平台执行可拖拽spark任务实际实行的是dolphinscheduler调度器中项目下工作流下的某一节点而已，不是执行整个工作流。

共用的依赖

<!--httpclient-->
<dependency>
    <groupId>commons-httpclient</groupId>
    <artifactId>commons-httpclient</artifactId>
    <version>3.1</version>
</dependency>

共用配置文件

dolphinscheduler.token=xxx
dolphinscheduler.address=http://IP:12345

共用代码

@Autowired
private RestTemplate restTemplate;
@Value("${dolphinscheduler.token}")
String token;
@Value("${dolphinscheduler.address}")
String address;
public static final int ZERO = 0;
public static final int SUCCESS = 200;
@Autowired
private DragSparkTaskService dragSparkTaskService;
@Value("${spark.main.class}")
String mainClass;
public static final String CREATE = "create";
public static final String UPDATE = "update";
public static final String ADD = "add";
public static final String DELETE = "delete";
public static final String ONLINE = "ONLINE";
public static final String OFFLINE = "OFFLINE";
public static final int ONE_THOUSAND_AND_FIVE_HUNDRED = 1500;
public static final int SIX = 6;
public static final int EIGHTY = 80;
public static final int THREE = 3;
@Autowired
private StringRedisTemplate redisTemplate;
@Value("${drag.task.state}")
String dragTaskState;
@Autowired
private DragSparkTaskMapper dragSparkTaskMapper;

1.执行任务

代码：

/**
     * 运行流程实例
     * @param projectName 项目名称
     * @param request request
     * @param dragSparkTaskId 任务ID
     * @author liudz
     * @date 2021/5/7
     * @return 执行结果
     **/
    @GetMapping("/project/process/start")
    public DolphinschedulerResponse startProcessInstance(
            @RequestParam("projectName") String projectName, @RequestParam("dragSparkTaskId") Integer dragSparkTaskId,
            HttpServletRequest request) {
   
        try {
   
            Long userId = Long.valueOf(request.getUserPrincipal().getName());
            DolphinschedulerResponse processInfoList = getUserProcess(projectName);
            if (processInfoList.getCode() != ZERO) {
   
                return processInfoList;
            }
            JSONObject processJson = new JSONObject();
            log.info("--(1)getUserProcess--success:{}", processInfoList);
            List<Map<String, Object>> list = (List<Map<String, Object>>) processInfoList.getData();
            for (Map<String, Object> map : list) {
   
                if (map.get("name").equals(userId + "-dragSparkTask")) {
   
                    processJson.fluentPutAll(map);
                }
            }
            if (processJson.getString(DictionaryEnum.RELEASE_STATE.getFiledString()).equals(OFFLINE)) {
   
                releaseProcessDefinition(projectName, userId + "-dragSparkTask",
                        processJson.getInteger("id"), 1);
                log.info("--(2)releaseProcessDefinition--ONLINE--success");
            }
            String postURL = address + "/dolphinscheduler/projects/" + URLEncoder.encode(projectName, "utf-8")
                   + "/executors/start-process-instance";
            PostMethod postMethod = new PostMethod(postURL);
            postMethod.setRequestHeader("Content-Type",
                    "application/x-www-form-urlencoded;charset=utf-8");
            postMethod.setRequestHeader("token", token);
            // 参数设置，需要注意的就是里边不能传NULL，要传空字符串
            NameValuePair[] data = packageNameValuePair(processJson, dragSparkTaskId);
            postMethod.setRequestBody(data);
            org.apache.commons.httpclient.HttpClient httpClient = new org.apache.commons.httpclient.HttpClient();
            httpClient.executeMethod(postMethod);
            JSONObject result = JSONObject.parseObject(postMethod.getResponseBodyAsString());
            log.info("--(2)startProcessInstance--result:{}", result);
            if (!result.get(DictionaryEnum.CODE.getFiledString()).equals(ZERO)) {
   
                return DolphinschedulerResponse.error(result.getInteger("code"), result.getString("msg"));
            }
            redisTemplate.opsForValue().set(dragTaskState + dragSparkTaskId, "1", 1, TimeUnit.HOURS);
            DragSparkTask drag = new DragSparkTask();
            drag.setId(Long.valueOf(dragSparkTaskId));
            drag.setState("1");
            drag.setCreateId(userId);
            dragSparkTaskMapper.updateDragSparkTask(drag);
            log.info("--(3)----updateDragSparkTask--success!");
        } catch (Exception e) {
   
            log.info("请求异常:{}", e);
        }
        return DolphinschedulerResponse.success();
    }

/**
     *  packageNameValuePair封装参数
     * @param processJson 工作流json
     * @param dragSparkTaskId 任务ID
     * @author liudz
     * @date 2021/5/14
     * @return NameValuePair
     **/
    public NameValuePair[] packageNameValuePair(JSONObject processJson, Integer dragSparkTaskId) {
   
        NameValuePair[] data = {
   
                new NameValuePair("failureStrategy", "CONTINUE"),
                new NameValuePair("processDefinitionId", processJson.getString("id")),
                new NameValuePair("processInstancePriority", "MEDIUM"),
                new NameValuePair("warningGroupId", "0"),
                new NameValuePair("warningType", "NONE"),
                new NameValuePair("runMode", "RUN_MODE_SERIAL"),
                new NameValuePair("startNodeList", "spark-" + dragSparkTaskId),
                new NameValuePair("taskDependType", "TASK_POST"),
                new NameValuePair("workerGroup", "default")};
        return data;
    }

/**
     *  解析节点和线，拼接nodesArray
     * @param jsonObject 模型task
     * @author liudz
     * @date 2020/12/10
     * @return 填充后的nodesArray
     **/
    public JSONArray parseLineAndNode(JSONObject jsonObject) {
   
        JSONArray edgesArray = jsonObject.getJSONArray("edges");
        JSONArray nodesArray = jsonObject.getJSONArray("nodes");
        for (int i = 0; i < edgesArray.size(); i++) {
   
            JSONObject edgeJson = edgesArray.getJSONObject(i);
            for (int j = 0; j < nodesArray.size(); j++) {
   
                JSONObject nodeJson = nodesArray.getJSONObject(j);
                String nodeSourceId = edgeJson.getString("source");
                String nodeTargetId = edgeJson.getString("target");
if ("breakUp".equals(nodeJson.getString("modelType")) && nodeSourceId.equals(nodeJson.getString("id"))) {
   
                    double fraction = edgeJson.getJSONObject("config").getDoubleValue("fraction");
                    if (fraction > Double.parseDouble("0.5")) {
   
                        nodeJson.getJSONObject("config").put("fraction", fraction);
                    }
                    if (nodeJson.getJSONArray("firstOutputs").size() == 0) {
   
                        String[] outputsArr = JavaTools.arrayInsert(nodeJson.getJSONArray("firstOutputs").
                                toArray(new String[nodeJson.getJSONArray("firstOutputs").size()]), nodeTargetId);
                        nodeJson.put("firstOutputs", outputsArr);
                        continue;
                    } else {
   
                        String[] outputsArr = JavaTools.arrayInsert(nodeJson.getJSONArray("secondOutputs").
                                toArray(new String[nodeJson.getJSONArray("secondOutputs").size()]), nodeTargetId);
                        nodeJson.put("secondOutputs", outputsArr);
                        continue;
                    }
                } else {
   
                    if (nodeSourceId.equals(nodeJson.getString("id"))) {
   
                        String[] outputsArr = JavaTools.arrayInsert(nodeJson.getJSONArray("outputs").
                                toArray(new String[nodeJson.getJSONArray("outputs").size()]), nodeTargetId);
                        nodeJson.put("outputs", outputsArr);
                        continue;
                    }
                    if (nodeTargetId.equals(nodeJson.getString("id"))) {
   
                        String[] inputsArr = JavaTools.arrayInsert(nodeJson.getJSONArray("inputs").
                                toArray(new String[nodeJson.getJSONArray("inputs").size()]), nodeSourceId);
                        nodeJson.put("inputs", inputsArr);
                        continue;
                    }
                }
            }
        }
        for (int j = 0; j < nodesArray.size(); j++) {
   
            JSONObject nodeJson = nodesArray.getJSONObject(j);
            String label = nodeJson.getString("modelType") + "_" + nodeJson.getString("id");
            String name = nodeJson.getString("label") + "_" + nodeJson.getString("id");
            nodeJson.put("name", name);
            nodeJson.put("label", label);
        }
        return nodesArray;
    }

2.停止任务

代码：

/**
     * stopProcessSparkTask
     * @param id id
     * @param executeType executeType
     * @param projectName 项目名称
     * @return Response
     * @author: liudz
     * @author: lty update 2020/5/27
     * @date: 2020/4/28 10:31
     */
    @GetMapping(value = "/project/execute/{projectName}/{id}/{executeType}")
    public Response<String> stopProcessSparkTask(@PathVariable("projectName") String projectName,
                                 @PathVariable("id") Long id, @PathVariable("executeType") String executeType) {
   
        log.info("--(1)stopProcessSparkTask--begin--projectName:{},id:{},executeType:{}", projectName, id, executeType);
        try {
   
            HttpHeaders headers = new HttpHeaders();
            headers.set("token", token);
            headers.set("Content-Type", "application/json");
            HttpEntity requestEntity = new HttpEntity(headers);
            ResponseEntity<JSONObject> response = restTemplate.exchange(address + "/"
   + "dolphinscheduler/projects/" + projectName + "/task-instance/list-paging?"
   + "pageNo=1&pageSize=100&taskName=spark-" + id, HttpMethod.GET, requestEntity, JSONObject.class);
List<Map<String, Object>> list = (List<Map<String, Object>>) response.getBody().getJSONObject("data").get("totalList");
            Integer processInstanceId = null;
            for (Map<String, Object> map : list) {
   
                if (map.get("state").equals("RUNNING_EXEUTION")) {
   
                    processInstanceId = Integer.valueOf(map.get("processInstanceId").toString());
                }
            }
            log.info("--(2)getProcessInstanceId--success--:{}", processInstanceId);
            String postURL = address + "/dolphinscheduler/projects/"
                    + URLEncoder.encode(projectName, "utf-8") + "/executors/execute";
            PostMethod postMethod = new PostMethod(postURL);
            postMethod.setRequestHeader("Content-Type",
                    "application/x-www-form-urlencoded;charset=utf-8");
            postMethod.setRequestHeader("token", token);
            NameValuePair[] data = {
   new NameValuePair("executeType", executeType),
                    new NameValuePair("processInstanceId", processInstanceId.toString())};
            postMethod.setRequestBody(data);
            org.apache.commons.httpclient.HttpClient httpClient = new org.apache.commons.httpclient.HttpClient();
            httpClient.executeMethod(postMethod);
            JSONObject result = JSONObject.parseObject(postMethod.getResponseBodyAsString());
            if (!result.get(DictionaryEnum.CODE.getFiledString()).equals(ZERO)) {
   
                return Response.error(result.getInteger("code"), result.getString("msg"));
            }
            log.info("--(3)stopProcessSparkTask--success--:{}", result);
            redisTemplate.opsForValue().set(dragTaskState + id, "0", 1, TimeUnit.HOURS);
            DragSparkTaskVo dragSparkTaskVo = new DragSparkTaskVo();
            dragSparkTaskVo.setId(id);
            dragSparkTaskVo.setState("0");
            dragSparkTaskService.updateDragSparkTask(dragSparkTaskVo);
            log.info("--(4)updateDragSparkTask--success");
        } catch (UnsupportedEncodingException e) {
   
            log.info("UnsupportedEncodingException:{}", e);
        } catch (HttpException e) {
   
            log.info("HttpException:{}", e);
        } catch (IOException e) {
   
            log.info("IOException:{}", e);
        }
        return Response.success();
    }

三、本人相关其他文章链接

1.springboot项目集成dolphinscheduler调度器可拖拽spark任务管理：
https://blog.csdn.net/a924382407/article/details/117119831

2.springboot项目集成dolphinscheduler调度器实现datax数据同步任务：
https://blog.csdn.net/a924382407/article/details/120951230

3.springboot项目集成dolphinscheduler调度器项目管理：
https://blog.csdn.net/a924382407/article/details/117118931

4.springboot项目集成大数据第三方dolphinscheduler调度器执行/停止任务
https://blog.csdn.net/a924382407/article/details/117121181

5.springboot项目集成大数据第三方dolphinscheduler调度器
https://blog.csdn.net/a924382407/article/details/117113848

springboot项目集成大数据第三方dolphinscheduler调度器执行/停止任务

一、功能清单

二、执行/停止任务

共用的依赖

共用配置文件

共用代码

1.执行任务

2.停止任务

三、本人相关其他文章链接

大数据与机器学习

热门文章

最新文章

相关产品

相关课程

相关电子书

springboot项目集成大数据第三方dolphinscheduler调度器 执行/停止任务

一、功能清单

二、执行/停止任务

共用的依赖

共用配置文件

共用代码

1.执行任务

2.停止任务

三、本人相关其他文章链接

大数据与机器学习

热门文章

最新文章

相关产品

相关课程

相关电子书

springboot项目集成大数据第三方dolphinscheduler调度器执行/停止任务