MapReduce源码分析之JobSubmitter(一)
JobSubmitter,顾名思义,它是MapReduce中作业提交者,而实际上JobSubmitter除了构造方法外,对外提供的唯一一个非private成员变量或方法就是submitJobInternal()方法,它是提交Job的内部方法,实现了提交Job的所有业务逻辑。本文,我们将深入研究MapReduce中用于提交Job的组件JobSubmitter。
首先,我们先看下JobSubmitter的类成员变量,如下:
- // 文件系统FileSystem实例
- private FileSystem jtFs;
- // 客户端通信协议ClientProtocol实例
- private ClientProtocol submitClient;
- // 提交作业的主机名
- private String submitHostName;
- // 提交作业的主机地址
- private String submitHostAddress;
它一共有四个类成员变量,分别为:
1、文件系统FileSystem实例jtFs:用于操作作业运行需要的各种文件等;
2、客户端通信协议ClientProtocol实例submitClient:用于与集群交互,完成作业提交、作业状态查询等;
3、提交作业的主机名submitHostName;
4、提交作业的主机地址submitHostAddress。
其中,客户端通信协议ClientProtocol实例submitClient是通过Cluster的客户端通信协议ClientProtocol实例client来赋值的,我们在《MapReduce源码分析之新API作业提交(二):连接集群》一文中曾经提到过,它根据MapReduce中参数mapreduce.framework.name的配置为yarn或local,有Yarn模式的YARNRunner和Local模式的LocalJobRunner两种情况。
接下来,我们再看下JobSubmitter的构造函数,如下:
- JobSubmitter(FileSystem submitFs, ClientProtocol submitClient)
- throws IOException {
- // 根据入参赋值成员变量submitClient、jtFs
- this.submitClient = submitClient;
- this.jtFs = submitFs;
- }
很简单,根据入参赋值成员变量submitClient、jtFs而已。
关键的来了,我们看下JobSubmitter唯一的对外核心功能方法submitJobInternal(),它被用于提交作业至集群,代码如下:
- /**
- * Internal method for submitting jobs to the system.
- *
- * <p>The job submission process involves:
- * <ol>
- * <li>
- * Checking the input and output specifications of the job.
- * </li>
- * <li>
- * Computing the {@link InputSplit}s for the job.
- * </li>
- * <li>
- * Setup the requisite accounting information for the
- * {@link DistributedCache} of the job, if necessary.
- * </li>
- * <li>
- * Copying the job's jar and configuration to the map-reduce system
- * directory on the distributed file-system.
- * </li>
- * <li>
- * Submitting the job to the <code>JobTracker</code> and optionally
- * monitoring it's status.
- * </li>
- * </ol></p>
- * @param job the configuration to submit
- * @param cluster the handle to the Cluster
- * @throws ClassNotFoundException
- * @throws InterruptedException
- * @throws IOException
- */
- JobStatus submitJobInternal(Job job, Cluster cluster)
- throws ClassNotFoundException, InterruptedException, IOException {
- //validate the jobs output specs
- // 调用checkSpecs()方法,校验作业输出路径是否配置,且是否已存在,
- // 正确的情况应该是已配置且未存在,输出路径配置参数为mapreduce.output.fileoutputformat.outputdir,
- // 之前WordCount作业的输出路径配置为hdfs://nameservice1/output/output
- checkSpecs(job);
- // 从作业job中获取配置信息conf
- Configuration conf = job.getConfiguration();
- // 调用addMRFrameworkToDistributedCache()方法添加应用框架路径到分布式缓存中
- addMRFrameworkToDistributedCache(conf);
- // 通过JobSubmissionFiles的getStagingDir()静态方法获取作业执行时阶段区域路径jobStagingArea
- // 取参数yarn.app.mapreduce.am.staging-dir,参数未配置默认为/tmp/hadoop-yarn/staging
- // 然后后面是/提交作业用户名/.staging
- // 通过之前的WordCount任务的执行,我们查看历史记录,得知参数yarn.app.mapreduce.am.staging-dir配置的为/user,
- // 而提交作业用户名为hdfs,所以完整的路径应该为/user/hdfs/.staging
- Path jobStagingArea = JobSubmissionFiles.getStagingDir(cluster, conf);
- //configure the command line options correctly on the submitting dfs
- // 获取当前本机地址
- InetAddress ip = InetAddress.getLocalHost();
- // 确定提交作业的主机地址、主机名,并设置入配置信息conf,对应参数分别为
- // mapreduce.job.submithostname
- // mapreduce.job.submithostaddress
- if (ip != null) {
- submitHostAddress = ip.getHostAddress();
- submitHostName = ip.getHostName();
- conf.set(MRJobConfig.JOB_SUBMITHOST,submitHostName);
- conf.set(MRJobConfig.JOB_SUBMITHOSTADDR,submitHostAddress);
- }
- // 生成作业ID,即JobID实例jobId
- JobID jobId = submitClient.getNewJobID();
- // 将jobId设置入job
- job.setJobID(jobId);
- // 构造提交作业路径Path实例submitJobDir,jobStagingArea后接/jobId,比如/job_1459913635503_0005
- // 之前WordCount作业的完整路径为/user/hdfs/.staging/job_1459913635503_0005
- Path submitJobDir = new Path(jobStagingArea, jobId.toString());
- JobStatus status = null;
- // 设置作业一些参数:
- try {
- // 设置mapreduce.job.user.name为当前用户,之前的WordCount示例配置的为hdfs用户
- conf.set(MRJobConfig.USER_NAME,
- UserGroupInformation.getCurrentUser().getShortUserName());
- // 设置hadoop.http.filter.initializers为AmFilterInitializer
- conf.set("hadoop.http.filter.initializers",
- "org.apache.hadoop.yarn.server.webproxy.amfilter.AmFilterInitializer");
- // 设置mapreduce.job.dir为submitJobDir,比如/user/hdfs/.staging/job_1459913635503_0005
- conf.set(MRJobConfig.MAPREDUCE_JOB_DIR, submitJobDir.toString());
- LOG.debug("Configuring job " + jobId + " with " + submitJobDir
- + " as the submit dir");
- // get delegation token for the dir
- // 获取路径的授权令牌:调用TokenCache的obtainTokensForNamenodes()静态方法
- TokenCache.obtainTokensForNamenodes(job.getCredentials(),
- new Path[] { submitJobDir }, conf);
- // 获取密钥和令牌,并将它们存储到令牌缓存TokenCache中
- populateTokenCache(conf, job.getCredentials());
- // generate a secret to authenticate shuffle transfers
- if (TokenCache.getShuffleSecretKey(job.getCredentials()) == null) {
- KeyGenerator keyGen;
- try {
- int keyLen = CryptoUtils.isShuffleEncrypted(conf)
- ? conf.getInt(MRJobConfig.MR_ENCRYPTED_INTERMEDIATE_DATA_KEY_SIZE_BITS,
- MRJobConfig.DEFAULT_MR_ENCRYPTED_INTERMEDIATE_DATA_KEY_SIZE_BITS)
- : SHUFFLE_KEY_LENGTH;
- keyGen = KeyGenerator.getInstance(SHUFFLE_KEYGEN_ALGORITHM);
- keyGen.init(keyLen);
- } catch (NoSuchAlgorithmException e) {
- throw new IOException("Error generating shuffle secret key", e);
- }
- SecretKey shuffleKey = keyGen.generateKey();
- TokenCache.setShuffleSecretKey(shuffleKey.getEncoded(),
- job.getCredentials());
- }
- // 复制并且配置相关文件
- copyAndConfigureFiles(job, submitJobDir);
- // 获取配置文件路径:job.xml
- Path submitJobFile = JobSubmissionFiles.getJobConfPath(submitJobDir);
- // Create the splits for the job
- LOG.debug("Creating splits at " + jtFs.makeQualified(submitJobDir));
- // 调用writeSplits()方法,写分片数据文件job.split和分片元数据文件job.splitmetainfo,
- // 并获得计算得到的map任务数目maps
- int maps = writeSplits(job, submitJobDir);
- // 配置信息中设置map任务数目mapreduce.job.maps为上面得到的maps
- conf.setInt(MRJobConfig.NUM_MAPS, maps);
- LOG.info("number of splits:" + maps);
- // write "queue admins of the queue to which job is being submitted"
- // to job file.
- // 获取作业队列名queue,取参数mapreduce.job.queuename,参数未配置默认为default,
- // 之前的WordCount任务示例中,作业队列名queue就为default
- String queue = conf.get(MRJobConfig.QUEUE_NAME,
- JobConf.DEFAULT_QUEUE_NAME);
- // 获取队列的访问权限控制列表AccessControlList实例acl,通过客户端通信协议ClientProtocol实例submitClient的getQueueAdmins()方法,传入队列名queue,
- // 实际上之前的WordCount任务示例中,这里获取的是*
- AccessControlList acl = submitClient.getQueueAdmins(queue);
- // 配置信息中设置队列参数mapred.queue.default.acl-administer-jobs
- // 之前的WordCount任务示例中,该参数被设置成为*
- conf.set(toFullPropertyName(queue,
- QueueACL.ADMINISTER_JOBS.getAclName()), acl.getAclString());
- // removing jobtoken referrals before copying the jobconf to HDFS
- // as the tasks don't need this setting, actually they may break
- // because of it if present as the referral will point to a
- // different job.
- // 清空缓存的令牌
- TokenCache.cleanUpTokenReferral(conf);
- // 根据参数确定是否需要追踪令牌ID
- // 取参数mapreduce.job.token.tracking.ids.enabled,参数未配置默认为false
- if (conf.getBoolean(
- MRJobConfig.JOB_TOKEN_TRACKING_IDS_ENABLED,
- MRJobConfig.DEFAULT_JOB_TOKEN_TRACKING_IDS_ENABLED)) {
- // 通过job获取令牌ID,并存储到trackingIds列表中
- // Add HDFS tracking ids
- ArrayList<String> trackingIds = new ArrayList<String>();
- for (Token<? extends TokenIdentifier> t :
- job.getCredentials().getAllTokens()) {
- trackingIds.add(t.decodeIdentifier().getTrackingId());
- }
- // 将trackingIds列表中的内容设置到参数mapreduce.job.token.tracking.ids中
- conf.setStrings(MRJobConfig.JOB_TOKEN_TRACKING_IDS,
- trackingIds.toArray(new String[trackingIds.size()]));
- }
- // Set reservation info if it exists
- // 如有必要,设置存在的预订信息
- // 参数为mapreduce.job.reservation.id
- ReservationId reservationId = job.getReservationId();
- if (reservationId != null) {
- conf.set(MRJobConfig.RESERVATION_ID, reservationId.toString());
- }
- // Write job file to submit dir
- // 调用writeConf()方法,写入作业配置信息至文件job.xml
- writeConf(conf, submitJobFile);
- //
- // Now, actually submit the job (using the submit name)
- // 调用printTokens()方法打印令牌信息到Log文件
- printTokens(jobId, job.getCredentials());
- // 通过客户端通信协议ClientProtocol实例submitClient的submitJob()方法提交作业,
- // 并获取作业状态JobStatus实例status
- // 由集群连接一文的分析我们可以知道,这个submitClient实际上是YARNRunner或LocalJobRunner对象,
- // 最终调用的是二者的submitJob()方法,我们留待以后分析
- status = submitClient.submitJob(
- jobId, submitJobDir.toString(), job.getCredentials());
- // 如果作业状态JobStatus实例status不为null,直接返回,否则抛出无法加载作业的IO异常
- if (status != null) {
- return status;
- } else {
- throw new IOException("Could not launch job");
- }
- } finally {
- // 最终,抛出无法加载作业的IO异常前,调用文件系统FileSystem实例jtFs的delete()方法,
- // 删除作业提交的相关目录或文件submitJobDir
- if (status == null) {
- LOG.info("Cleaning up the staging area " + submitJobDir);
- if (jtFs != null && submitJobDir != null)
- jtFs.delete(submitJobDir, true);
- }
- }
- }
submitJobInternal()方法篇幅比较长,逻辑也很复杂,本文先介绍下它的大体逻辑,后续分文会介绍各个环节的详细内容,且下面涉及到的之前WordCount作业示例在《Hadoop2.6.0版本MapReudce示例之WordCount(一)》及其姊妹篇中,敬请注意!submitJobInternal()方法大体逻辑如下:
1、调用checkSpecs()方法,校验作业输出路径是否配置,且是否已存在:
正确的情况应该是已配置且未存在,输出路径配置参数为mapreduce.output.fileoutputformat.outputdir,之前WordCount作业的输出路径配置为hdfs://nameservice1/output/output;
2、从作业job中获取配置信息conf;
3、调用addMRFrameworkToDistributedCache()方法添加应用框架路径到分布式缓存中;
4、通过JobSubmissionFiles的getStagingDir()静态方法获取作业执行时阶段区域路径jobStagingArea:
取参数yarn.app.mapreduce.am.staging-dir,参数未配置默认为/tmp/Hadoop-yarn/staging,然后后面是/提交作业用户名/.staging,通过之前的WordCount任务的执行,我们查看历史记录,得知参数yarn.app.mapreduce.am.staging-dir配置的为/user,而提交作业用户名为hdfs,所以完整的路径应该为/user/hdfs/.staging;
5、获取当前本机地址ip;
6、确定提交作业的主机地址、主机名,并设置入配置信息conf,对应参数分别为mapreduce.job.submithostname、mapreduce.job.submithostaddress;
7、生成作业ID,即JobID实例jobId:
通过客户端通信协议ClientProtocol实例submitClient的getNewJobID()方法生成作业ID,即JobID实例jobId;
8、 将jobId设置入job;
9、构造提交作业路径Path实例submitJobDir:
jobStagingArea后接/jobId,比如/job_1459913635503_0005,之前WordCount作业的完整路径为/user/hdfs/.staging/job_1459913635503_0005;
10、设置作业一些参数:
10.1、设置mapreduce.job.user.name为当前用户,之前的WordCount示例配置的为hdfs用户;
10.2、设置hadoop.http.filter.initializers为AmFilterInitializer;
10.3、设置mapreduce.job.dir为submitJobDir,比如/user/hdfs/.staging/job_1459913635503_0005;
11、获取路径的授权令牌:调用TokenCache的obtainTokensForNamenodes()静态方法;
12、通过populateTokenCache()方法获取密钥和令牌,并将它们存储到令牌缓存TokenCache中;
14、复制并且配置相关文件:通过copyAndConfigureFiles()方法实现;
15、获取配置文件路径:job.xml;
16、调用writeSplits()方法,写分片数据文件job.split和分片元数据文件job.splitmetainfo,并获得计算得到的map任务数目maps;
17、配置信息中设置map任务数目mapreduce.job.maps为上面得到的maps;
18、获取作业队列名queue,取参数mapreduce.job.queuename,参数未配置默认为default,之前的WordCount任务示例中,作业队列名queue就为default;
19、获取队列的访问权限控制列表AccessControlList实例acl:
通过客户端通信协议ClientProtocol实例submitClient的getQueueAdmins()方法,传入队列名queue,实际上之前的WordCount任务示例中,这里获取的是*;
20、配置信息中设置队列参数mapred.queue.default.acl-administer-jobs,之前的WordCount任务示例中,该参数被设置成为*;
21、清空缓存的令牌:通过TokenCache的cleanUpTokenReferral()方法实现;
22、根据参数确定是否需要追踪令牌ID,如果需要的话:
取参数mapreduce.job.token.tracking.ids.enabled,参数未配置默认为false,通过job获取令牌ID,并存储到trackingIds列表中,将trackingIds列表中的内容设置到参数mapreduce.job.token.tracking.ids中;
23、如有必要,设置存在的预订信息:参数为mapreduce.job.reservation.id;
24、调用writeConf()方法,写入作业配置信息至文件job.xml;
25、调用printTokens()方法打印令牌信息到Log文件;
26、通过客户端通信协议ClientProtocol实例submitClient的submitJob()方法提交作业,并获取作业状态JobStatus实例status:
由集群连接一文的分析我们可以知道,这个submitClient实际上是YARNRunner或LocalJobRunner对象,最终调用的是二者的submitJob()方法,我们留待以后分析;
27、如果作业状态JobStatus实例status不为null,直接返回,否则抛出无法加载作业的IO异常:
最终,抛出无法加载作业的IO异常前,调用文件系统FileSystem实例jtFs的delete()方法,删除作业提交的相关目录或文件submitJobDir。
整体流程如上,对于关键步骤的主要细节,限于篇幅,敬请关注《MapReduce源码分析之JobSubmitter(二)》!