分布式调度框架。
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 

286 lines
14 KiB

/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.dolphinscheduler.server.master.service;
import org.apache.dolphinscheduler.common.constants.Constants;
import org.apache.dolphinscheduler.common.enums.Flag;
import org.apache.dolphinscheduler.common.enums.StateEventType;
import org.apache.dolphinscheduler.common.model.Server;
import org.apache.dolphinscheduler.dao.entity.ProcessInstance;
import org.apache.dolphinscheduler.dao.entity.TaskInstance;
import org.apache.dolphinscheduler.dao.repository.TaskInstanceDao;
import org.apache.dolphinscheduler.extract.base.client.SingletonJdkDynamicRpcClientProxyFactory;
import org.apache.dolphinscheduler.extract.worker.IWorkerLogService;
import org.apache.dolphinscheduler.extract.worker.transportor.GetAppIdRequest;
import org.apache.dolphinscheduler.extract.worker.transportor.GetAppIdResponse;
import org.apache.dolphinscheduler.plugin.task.api.TaskExecutionContext;
import org.apache.dolphinscheduler.plugin.task.api.enums.TaskExecutionStatus;
import org.apache.dolphinscheduler.plugin.task.api.utils.LogUtils;
import org.apache.dolphinscheduler.registry.api.RegistryClient;
import org.apache.dolphinscheduler.registry.api.enums.RegistryNodeType;
import org.apache.dolphinscheduler.server.master.builder.TaskExecutionContextBuilder;
import org.apache.dolphinscheduler.server.master.cache.ProcessInstanceExecCacheManager;
import org.apache.dolphinscheduler.server.master.config.MasterConfig;
import org.apache.dolphinscheduler.server.master.event.TaskStateEvent;
import org.apache.dolphinscheduler.server.master.metrics.TaskMetrics;
import org.apache.dolphinscheduler.server.master.runner.WorkflowExecuteRunnable;
import org.apache.dolphinscheduler.server.master.runner.WorkflowExecuteThreadPool;
import org.apache.dolphinscheduler.server.master.utils.TaskUtils;
import org.apache.dolphinscheduler.service.process.ProcessService;
import org.apache.dolphinscheduler.service.utils.ProcessUtils;
import org.apache.commons.collections4.CollectionUtils;
import org.apache.commons.lang3.StringUtils;
import org.apache.commons.lang3.time.StopWatch;
import java.util.Date;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Optional;
import java.util.concurrent.TimeUnit;
import java.util.stream.Collectors;
import javax.annotation.Nullable;
import lombok.NonNull;
import lombok.extern.slf4j.Slf4j;
import org.springframework.stereotype.Service;
@Service
@Slf4j
public class WorkerFailoverService {
private final RegistryClient registryClient;
private final MasterConfig masterConfig;
private final ProcessService processService;
private final WorkflowExecuteThreadPool workflowExecuteThreadPool;
private final ProcessInstanceExecCacheManager cacheManager;
private final String localAddress;
private final TaskInstanceDao taskInstanceDao;
public WorkerFailoverService(@NonNull RegistryClient registryClient,
@NonNull MasterConfig masterConfig,
@NonNull ProcessService processService,
@NonNull WorkflowExecuteThreadPool workflowExecuteThreadPool,
@NonNull ProcessInstanceExecCacheManager cacheManager,
@NonNull TaskInstanceDao taskInstanceDao) {
this.registryClient = registryClient;
this.masterConfig = masterConfig;
this.processService = processService;
this.workflowExecuteThreadPool = workflowExecuteThreadPool;
this.cacheManager = cacheManager;
this.localAddress = masterConfig.getMasterAddress();
this.taskInstanceDao = taskInstanceDao;
}
/**
* Do the worker failover. Will find the SUBMITTED_SUCCESS/DISPATCH/RUNNING_EXECUTION/DELAY_EXECUTION/READY_PAUSE/READY_STOP tasks belong the given worker,
* and failover these tasks.
* <p>
* Note: When we do worker failover, the master will only failover the processInstance belongs to the current master.
*
* @param workerHost worker host
*/
public void failoverWorker(@NonNull String workerHost) {
log.info("Worker[{}] failover starting", workerHost);
final StopWatch failoverTimeCost = StopWatch.createStarted();
// we query the task instance from cache, so that we can directly update the cache
final Optional<Date> needFailoverWorkerStartTime =
getServerStartupTime(registryClient.getServerList(RegistryNodeType.WORKER), workerHost);
final List<TaskInstance> needFailoverTaskInstanceList = getNeedFailoverTaskInstance(workerHost);
if (CollectionUtils.isEmpty(needFailoverTaskInstanceList)) {
log.info("Worker[{}] failover finished there are no taskInstance need to failover", workerHost);
return;
}
log.info(
"Worker[{}] failover there are {} taskInstance may need to failover, will do a deep check, taskInstanceIds: {}",
workerHost,
needFailoverTaskInstanceList.size(),
needFailoverTaskInstanceList.stream().map(TaskInstance::getId).collect(Collectors.toList()));
final Map<Integer, ProcessInstance> processInstanceCacheMap = new HashMap<>();
for (TaskInstance taskInstance : needFailoverTaskInstanceList) {
try {
LogUtils.setWorkflowAndTaskInstanceIDMDC(taskInstance.getProcessInstanceId(), taskInstance.getId());
ProcessInstance processInstance = processInstanceCacheMap.computeIfAbsent(
taskInstance.getProcessInstanceId(), k -> {
WorkflowExecuteRunnable workflowExecuteRunnable = cacheManager.getByProcessInstanceId(
taskInstance.getProcessInstanceId());
if (workflowExecuteRunnable == null) {
return null;
}
return workflowExecuteRunnable.getWorkflowExecuteContext()
.getWorkflowInstance();
});
if (!checkTaskInstanceNeedFailover(needFailoverWorkerStartTime, processInstance, taskInstance)) {
log.info("Worker[{}] the current taskInstance doesn't need to failover", workerHost);
continue;
}
log.info(
"Worker[{}] failover: begin to failover taskInstance, will set the status to NEED_FAULT_TOLERANCE",
workerHost);
failoverTaskInstance(processInstance, taskInstance);
log.info("Worker[{}] failover: Finish failover taskInstance", workerHost);
} catch (Exception ex) {
log.info("Worker[{}] failover taskInstance occur exception", workerHost, ex);
} finally {
LogUtils.removeWorkflowAndTaskInstanceIdMDC();
}
}
failoverTimeCost.stop();
log.info("Worker[{}] failover finished, useTime:{}ms",
workerHost,
failoverTimeCost.getTime(TimeUnit.MILLISECONDS));
}
/**
* failover task instance
* <p>
* 1. kill yarn/k8s job if run on worker and there are yarn/k8s jobs in tasks.
* 2. change task state from running to need failover.
* 3. try to notify local master
*
* @param processInstance
* @param taskInstance
*/
private void failoverTaskInstance(@NonNull ProcessInstance processInstance, @NonNull TaskInstance taskInstance) {
TaskMetrics.incTaskInstanceByState("failover");
taskInstance.setProcessInstance(processInstance);
if (!TaskUtils.isMasterTask(taskInstance.getTaskType())) {
killYarnTask(taskInstance, processInstance);
} else {
log.info("The failover taskInstance is a master task, no need to failover in worker failover");
}
taskInstance.setState(TaskExecutionStatus.NEED_FAULT_TOLERANCE);
taskInstance.setFlag(Flag.NO);
taskInstanceDao.upsertTaskInstance(taskInstance);
TaskStateEvent stateEvent = TaskStateEvent.builder()
.processInstanceId(processInstance.getId())
.taskInstanceId(taskInstance.getId())
.status(TaskExecutionStatus.NEED_FAULT_TOLERANCE)
.type(StateEventType.TASK_STATE_CHANGE)
.build();
workflowExecuteThreadPool.submitStateEvent(stateEvent);
}
/**
* task needs failover if task start before server starts
*
* @return true if task instance need fail over
*/
private boolean checkTaskInstanceNeedFailover(Optional<Date> needFailoverWorkerStartTime,
@Nullable ProcessInstance processInstance,
TaskInstance taskInstance) {
if (processInstance == null) {
log.error(
"Failover task instance error, cannot find the related processInstance form memory, this case shouldn't happened");
return false;
}
if (taskInstance == null) {
log.error("Master failover task instance error, taskInstance is null, this case shouldn't happened");
return false;
}
// only failover the task owned myself if worker down.
if (!StringUtils.equalsIgnoreCase(processInstance.getHost(), localAddress)) {
log.error(
"Master failover task instance error, the taskInstance's processInstance's host: {} is not the current master: {}",
processInstance.getHost(),
localAddress);
return false;
}
if (taskInstance.getState() != null && taskInstance.getState().isFinished()) {
// The taskInstance is already finished, doesn't need to failover
log.info("The task is already finished, doesn't need to failover");
return false;
}
if (!needFailoverWorkerStartTime.isPresent()) {
// The worker is still down
return true;
}
// The worker is active, may already send some new task to it
if (taskInstance.getSubmitTime() != null && taskInstance.getSubmitTime()
.after(needFailoverWorkerStartTime.get())) {
log.info(
"The taskInstance's submitTime: {} is after the need failover worker's start time: {}, the taskInstance is newly submit, it doesn't need to failover",
taskInstance.getSubmitTime(),
needFailoverWorkerStartTime.get());
return false;
}
return true;
}
private List<TaskInstance> getNeedFailoverTaskInstance(@NonNull String failoverWorkerHost) {
// we query the task instance from cache, so that we can directly update the cache
return cacheManager.getAll()
.stream()
.flatMap(workflowExecuteRunnable -> workflowExecuteRunnable.getAllTaskInstances().stream())
// If the worker is in dispatching and the host is not set
.filter(taskInstance -> failoverWorkerHost.equals(taskInstance.getHost())
&& taskInstance.getState().shouldFailover())
.collect(Collectors.toList());
}
private Optional<Date> getServerStartupTime(List<Server> servers, String host) {
if (CollectionUtils.isEmpty(servers)) {
return Optional.empty();
}
Date serverStartupTime = null;
for (Server server : servers) {
if (host.equals(server.getHost() + Constants.COLON + server.getPort())) {
serverStartupTime = server.getCreateTime();
break;
}
}
return Optional.ofNullable(serverStartupTime);
}
private void killYarnTask(TaskInstance taskInstance, ProcessInstance processInstance) {
try {
if (!masterConfig.isKillApplicationWhenTaskFailover()) {
return;
}
if (StringUtils.isEmpty(taskInstance.getHost()) || StringUtils.isEmpty(taskInstance.getLogPath())) {
return;
}
TaskExecutionContext taskExecutionContext = TaskExecutionContextBuilder.get()
.buildWorkflowInstanceHost(masterConfig.getMasterAddress())
.buildTaskInstanceRelatedInfo(taskInstance)
.buildProcessInstanceRelatedInfo(processInstance)
.buildProcessDefinitionRelatedInfo(processInstance.getProcessDefinition())
.create();
// only kill yarn/k8s job if exists , the local thread has exited
log.info("TaskInstance failover begin kill the task related yarn or k8s job");
IWorkerLogService iWorkerLogService = SingletonJdkDynamicRpcClientProxyFactory
.getProxyClient(taskInstance.getHost(), IWorkerLogService.class);
GetAppIdResponse getAppIdResponse =
iWorkerLogService.getAppId(new GetAppIdRequest(taskInstance.getId(), taskInstance.getLogPath()));
ProcessUtils.killApplication(getAppIdResponse.getAppIds(), taskExecutionContext);
} catch (Exception ex) {
log.error("Kill yarn task error", ex);
}
}
}