Yarn任务提交流程(源码分析)
关键词:yarn rm mapreduce 提交
Based on Hadoop 2.7.1
JobSubmitter
- addMRFrameworkToDistributedCache(Configuration conf) : mapreduce.application.framework.path, 用于指定其他framework的hdfs 路径配置,默认yarn的可以不管
- Token相关的方法:读取认证信息(支持二进制、json),并将其添加至相应的fileSystem中,以便以同样权限访问文件系统
- copyAndConfigureFiles(Job job, Path jobSubmitDir): 上传配置、jar、files、libjars、archives等
- submitJobInternal: 真正的提交任务接口
核心代码提交链
- JobSubmitter ->
- ClientProtocol(YARNRunner) ->
- ResourceMgrDelegate ->
- YarnClient(YarnClientImpl).submitApplication( ApplicationSubmissionContext appContext) ->
- 【RM】ApplicationClientProtocol(ClientRMService).submitApplication( SubmitApplicationRequest request) -> // fill ASC with dft values
RMAppManager.submitApplication( ApplicationSubmissionContext submissionContext, long submitTime, String user) ->
ApplicationSubmissionContext提交上下文,包含application各种元信息SubmitApplicationRequest提交Request对象
// Dispatcher is not yet started at this time, so these START events
// enqueued should be guaranteed to be first processed when dispatcher
// gets started.
this.rmContext.getDispatcher().getEventHandler()
.handle(new RMAppEvent(applicationId, RMAppEventType.START));
START -> APP_NEW_SAVED
stateMachineFactory.addTransition(RMAppState.NEW, RMAppState.NEW_SAVING,
RMAppEventType.START, new RMAppNewlySavingTransition())
//...
private static final class RMAppNewlySavingTransition extends RMAppTransition {
@Override
public void transition(RMAppImpl app, RMAppEvent event) {
// If recovery is enabled then store the application information in a
// non-blocking call so make sure that RM has stored the information
// needed to restart the AM after RM restart without further client
// communication
LOG.info("Storing application with id " + app.applicationId);
app.rmContext.getStateStore().storeNewApplication(app);
}
}
public synchronized void storeNewApplication(RMApp app) {
ApplicationSubmissionContext context = app
.getApplicationSubmissionContext();
assert context instanceof ApplicationSubmissionContextPBImpl;
ApplicationStateData appState =
ApplicationStateData.newInstance(
app.getSubmitTime(), app.getStartTime(), context, app.getUser());
dispatcher.getEventHandler().handle(new RMStateStoreAppEvent(appState));
}
private static final class AddApplicationToSchedulerTransition extends
RMAppTransition {
@Override
public void transition(RMAppImpl app, RMAppEvent event) {
app.handler.handle(new AppAddedSchedulerEvent(app.applicationId,
app.submissionContext.getQueue(), app.user,
app.submissionContext.getReservationID()));
}
}
AppAddedSchedulerEvent 会由配置的Scheduler来handle。
P.S. 看 event 部分代码的方法,
- 找出状态,比如 APP_NEW_SAVED,
- 找出handle这个状态的事件类
- 找出处理这个事件的具体逻辑 (这里可能逻辑最复杂)
- 找下一个事件
- 重复。。
ApplicationMaster
START -> APPNEWSAVED -> APP_ACCEPTED ....
后面是一些attempt的启动等各种事件的反复。直接跳到 AM 部分。
ResourceManager内有 createApplicationMasterLauncher() 和 createApplicationMasterService()
private void launch() throws IOException, YarnException {
connect();
ContainerId masterContainerID = masterContainer.getId();
ApplicationSubmissionContext applicationContext =
application.getSubmissionContext();
LOG.info("Setting up container " + masterContainer
+ " for AM " + application.getAppAttemptId());
ContainerLaunchContext launchContext =
createAMContainerLaunchContext(applicationContext, masterContainerID);
StartContainerRequest scRequest =
StartContainerRequest.newInstance(launchContext,
masterContainer.getContainerToken());
List<StartContainerRequest> list = new ArrayList<StartContainerRequest>();
list.add(scRequest);
StartContainersRequest allRequests =
StartContainersRequest.newInstance(list);
StartContainersResponse response =
containerMgrProxy.startContainers(allRequests);
if (response.getFailedRequests() != null
&& response.getFailedRequests().containsKey(masterContainerID)) {
Throwable t =
response.getFailedRequests().get(masterContainerID).deSerialize();
parseAndThrowException(t);
} else {
LOG.info("Done launching container " + masterContainer + " for AM "
+ application.getAppAttemptId());
}
}
private ContainerLaunchContext createAMContainerLaunchContext(
ApplicationSubmissionContext applicationMasterContext,
ContainerId containerID) throws IOException {
// Construct the actual Container
ContainerLaunchContext container =
applicationMasterContext.getAMContainerSpec();
LOG.info("Command to launch container "
+ containerID
+ " : "
+ StringUtils.arrayToString(container.getCommands().toArray(
new String[0])));
// Finalize the container
setupTokens(container, containerID);
return container;
}
注意以上其中两行:
- ContainerLaunchContext launchContext = createAMContainerLaunchContext(applicationContext, masterContainerID) 创建 AM 请求
- StartContainersResponse response = containerMgrProxy.startContainers(allRequests); 启动AM的容器并在容器内启动AM。
@Override
public ContainerLaunchContext getAMContainerSpec() {
ApplicationSubmissionContextProtoOrBuilder p = viaProto ? proto : builder;
if (this.amContainer != null) {
return amContainer;
} // Else via proto
if (!p.hasAmContainerSpec()) {
return null;
}
amContainer = convertFromProtoFormat(p.getAmContainerSpec());
return amContainer;
}
public class ApplicationSubmissionContextPBImpl
extends ApplicationSubmissionContext {
ApplicationSubmissionContextProto proto =
ApplicationSubmissionContextProto.getDefaultInstance();
ApplicationSubmissionContextProto.Builder builder = null;
boolean viaProto = false;
private ApplicationId applicationId = null;
private Priority priority = null;
private ContainerLaunchContext amContainer = null;
private Resource resource = null;
private Set<String> applicationTags = null;
private ResourceRequest amResourceRequest = null;
private LogAggregationContext logAggregationContext = null;
private ReservationId reservationId = null;
/// ...
}
接下来便是启动后的AppMaster 创建job,并通过AMRMClient向ResourceManager申请资源等。
Yarn任务提交流程(源码分析)的更多相关文章
- MapReduce之提交job源码分析 FileInputFormat源码解析
MapReduce之提交job源码分析 job 提交流程源码详解 //runner 类中提交job waitForCompletion() submit(); // 1 建立连接 connect(); ...
- Django rest framework 的认证流程(源码分析)
一.基本流程举例: urlpatterns = [ url(r'^admin/', admin.site.urls), url(r'^users/', views.HostView.as_view() ...
- springMvc的执行流程(源码分析)
1.在springMvc中负责处理请求的类为DispatcherServlet,这个类与我们传统的Servlet是一样的.我们来看看它的继承图 2. 我们发现DispatcherServlet也继承了 ...
- Spring Securtiy 认证流程(源码分析)
当用 Spring Security 框架进行认证时,你可能会遇到这样的问题: 你输入的用户名或密码不管是空还是错误,它的错误信息都是 Bad credentials. 那么如果你想根据不同的情况给出 ...
- Flink命令行提交job (源码分析)
这篇文章主要介绍从命令行到任务在Driver端运行的过程 通过flink run 命令提交jar包运行程序 以yarn 模式提交任务命令类似于: flink run -m yarn-cluster X ...
- hbase0.96 put流程 源码分析
无意间多瞄了一眼hbase0.98的代码,想复习下put流程.发现htable里面已经找不到processBatchOfPuts()奇怪了.看了半天原来变化还真大事实上0.96就没这个了,于是又搞了个 ...
- MapReduce——客户端提交任务源码分析
计算向数据移动 MR程序并不会在客户端执行任何的计算操作,它是为计算工作做好准备,例如计算出切片信息,直接影响到Map任务的并行度. 在Driver中提交任务时,会写到这样的语句: boolean r ...
- springmvc执行流程 源码分析
进入DispatcherServlet 执行onRefresh,然后执行初始化方法initStrategies.然后调用doService——>doDispatch. 根据继承关系执行Servl ...
- solr源码分析之solrclound
一.简介 SolrCloud是Solr4.0版本以后基于Solr和Zookeeper的分布式搜索方案.SolrCloud是Solr的基于Zookeeper一种部署方式.Solr可以以多种方式部署,例如 ...
- drf的基本使用、APIView源码分析和CBV源码拓展
cbv源码拓展 扩展,如果我在Book视图类中重写dispatch方法 -可以实现,在get,post方法执行之前或者之后执行代码,完成类似装饰器的效果 def dispatch(self, requ ...
随机推荐
- Spark Standalone spark-env.sh
export JAVA_HOME=/app/jdk export SPARK_MASTER_PORT=7077 export SPARK_MASTER_WEBUI_PORT=8080 export S ...
- SpringBoot系列: Spring项目异常处理最佳实践
===================================自定义异常类===================================稍具规模的项目, 一般都要自定义一组异常类, 这 ...
- visual studio code运行时报错,Cannot find module 'webpack'
前言 今天运行是visual studio code时,报了一个错误Cannot find module 'webpack' ,网上找了很多方法都没解决.下面一起来看看怎么解决 正文 报错: npm ...
- 两个Html页面之间值得传递
传值的页面:<a href='stockProductInfo.html?prdId="+v.prdID+"' target='_blank'></html> ...
- Java同步容器
一.为什么会出现同步容器 Java的集合框架中,主要有四大类别:List,Set,Queue,Map List,Set,Queue接口分别继承了Collection接口,Map本身是一个接口. 注意C ...
- Django-3-Template模板
模板是html文档+Django逻辑语句的组合. 一.变量和标签 变量通过{{ }}来表示,两个大括号中间是变量名. 标签通过{% %}来表示,就是Python中的函数和方法. 常用标签: {% fo ...
- Linux环境下安装SQL Server 2017
参考链接 https://docs.microsoft.com/zh-cn/sql/linux/quickstart-install-connect-red-hat?view=sql-server-2 ...
- ROS-PCL点云库的安装
网上各种说安装点云库还慢,中间出现一些没有依赖库的情况,但在我这里都是很顺利,难道是一开始安装了ROS的原因?不是很清楚,暂且将过程写下来. 一,下载PCL $ git clone https://g ...
- linux日志过滤某时间段的日志
sed -n '/2019-01-10 16:00*/,/2019-01-10 18:48*/p' nohup.out > 111.log
- iOS -- Effective Objective-C 阅读笔记 (9)
// 将类的实现方法代码反三到便于管理的数个分类之中. // 类中经常容易填满各种方法, 而这些方法的代码则全部堆在一个巨大的实现文件中, 有时这么做事不合理的,因为即使通过重构把这个类 ...