Beispiel #1
0
 private String copyFromLocal(FileSystem fs, Path basePath, String[] files) throws IOException {
   StringBuilder csv = new StringBuilder(files.length * (basePath.toString().length() + 16));
   for (String localFile : files) {
     Path src = new Path(localFile);
     String filename = src.getName();
     Path dst = new Path(basePath, filename);
     URI localFileURI = null;
     try {
       localFileURI = new URI(localFile);
     } catch (URISyntaxException e) {
       throw new IOException(e);
     }
     if (localFileURI.getScheme() == null || localFileURI.getScheme().startsWith("file")) {
       LOG.info("Copy {} from local filesystem to {}", localFile, dst);
       fs.copyFromLocalFile(false, true, src, dst);
     } else {
       LOG.info("Copy {} from DFS to {}", localFile, dst);
       FileUtil.copy(fs, src, fs, dst, false, true, conf);
     }
     if (csv.length() > 0) {
       csv.append(LIB_JARS_SEP);
     }
     csv.append(dst.toString());
   }
   return csv.toString();
 }
Beispiel #2
0
  /**
   * Launch application for the dag represented by this client.
   *
   * @throws YarnException
   * @throws IOException
   */
  public void startApplication() throws YarnException, IOException {
    Class<?>[] defaultClasses;

    if (applicationType.equals(YARN_APPLICATION_TYPE)) {
      // TODO restrict the security check to only check if security is enabled for webservices.
      if (UserGroupInformation.isSecurityEnabled()) {
        defaultClasses = DATATORRENT_SECURITY_CLASSES;
      } else {
        defaultClasses = DATATORRENT_CLASSES;
      }
    } else {
      throw new IllegalStateException(applicationType + " is not a valid application type.");
    }

    LinkedHashSet<String> localJarFiles = findJars(dag, defaultClasses);

    if (resources != null) {
      localJarFiles.addAll(resources);
    }

    YarnClusterMetrics clusterMetrics = yarnClient.getYarnClusterMetrics();
    LOG.info(
        "Got Cluster metric info from ASM"
            + ", numNodeManagers="
            + clusterMetrics.getNumNodeManagers());

    // GetClusterNodesRequest clusterNodesReq = Records.newRecord(GetClusterNodesRequest.class);
    // GetClusterNodesResponse clusterNodesResp =
    // rmClient.clientRM.getClusterNodes(clusterNodesReq);
    // LOG.info("Got Cluster node info from ASM");
    // for (NodeReport node : clusterNodesResp.getNodeReports()) {
    //  LOG.info("Got node report from ASM for"
    //           + ", nodeId=" + node.getNodeId()
    //           + ", nodeAddress" + node.getHttpAddress()
    //           + ", nodeRackName" + node.getRackName()
    //           + ", nodeNumContainers" + node.getNumContainers()
    //           + ", nodeHealthStatus" + node.getHealthReport());
    // }
    List<QueueUserACLInfo> listAclInfo = yarnClient.getQueueAclsInfo();
    for (QueueUserACLInfo aclInfo : listAclInfo) {
      for (QueueACL userAcl : aclInfo.getUserAcls()) {
        LOG.info(
            "User ACL Info for Queue"
                + ", queueName="
                + aclInfo.getQueueName()
                + ", userAcl="
                + userAcl.name());
      }
    }

    // Get a new application id
    YarnClientApplication newApp = yarnClient.createApplication();
    appId = newApp.getNewApplicationResponse().getApplicationId();

    // Dump out information about cluster capability as seen by the resource manager
    int maxMem = newApp.getNewApplicationResponse().getMaximumResourceCapability().getMemory();
    LOG.info("Max mem capabililty of resources in this cluster " + maxMem);
    int amMemory = dag.getMasterMemoryMB();
    if (amMemory > maxMem) {
      LOG.info(
          "AM memory specified above max threshold of cluster. Using max value."
              + ", specified="
              + amMemory
              + ", max="
              + maxMem);
      amMemory = maxMem;
    }

    if (dag.getAttributes().get(LogicalPlan.APPLICATION_ID) == null) {
      dag.setAttribute(LogicalPlan.APPLICATION_ID, appId.toString());
    }

    // Create launch context for app master
    LOG.info("Setting up application submission context for ASM");
    ApplicationSubmissionContext appContext = Records.newRecord(ApplicationSubmissionContext.class);

    // set the application id
    appContext.setApplicationId(appId);
    // set the application name
    appContext.setApplicationName(dag.getValue(LogicalPlan.APPLICATION_NAME));
    appContext.setApplicationType(this.applicationType);
    if (YARN_APPLICATION_TYPE.equals(this.applicationType)) {
      // appContext.setMaxAppAttempts(1); // no retries until Stram is HA
    }

    // Set up the container launch context for the application master
    ContainerLaunchContext amContainer = Records.newRecord(ContainerLaunchContext.class);

    // Setup security tokens
    // If security is enabled get ResourceManager and NameNode delegation tokens.
    // Set these tokens on the container so that they are sent as part of application submission.
    // This also sets them up for renewal by ResourceManager. The NameNode delegation rmToken
    // is also used by ResourceManager to fetch the jars from HDFS and set them up for the
    // application master launch.
    if (UserGroupInformation.isSecurityEnabled()) {
      Credentials credentials = new Credentials();
      String tokenRenewer = conf.get(YarnConfiguration.RM_PRINCIPAL);
      if (tokenRenewer == null || tokenRenewer.length() == 0) {
        throw new IOException("Can't get Master Kerberos principal for the RM to use as renewer");
      }

      // For now, only getting tokens for the default file-system.
      FileSystem fs = StramClientUtils.newFileSystemInstance(conf);
      try {
        final Token<?> tokens[] = fs.addDelegationTokens(tokenRenewer, credentials);
        if (tokens != null) {
          for (Token<?> token : tokens) {
            LOG.info("Got dt for " + fs.getUri() + "; " + token);
          }
        }
      } finally {
        fs.close();
      }

      addRMDelegationToken(tokenRenewer, credentials);

      DataOutputBuffer dob = new DataOutputBuffer();
      credentials.writeTokenStorageToStream(dob);
      ByteBuffer fsTokens = ByteBuffer.wrap(dob.getData(), 0, dob.getLength());
      amContainer.setTokens(fsTokens);
    }

    // set local resources for the application master
    // local files or archives as needed
    // In this scenario, the jar file for the application master is part of the local resources
    Map<String, LocalResource> localResources = new HashMap<String, LocalResource>();

    // copy required jar files to dfs, to be localized for containers
    FileSystem fs = StramClientUtils.newFileSystemInstance(conf);
    try {
      Path appsBasePath =
          new Path(StramClientUtils.getDTDFSRootDir(fs, conf), StramClientUtils.SUBDIR_APPS);
      Path appPath = new Path(appsBasePath, appId.toString());

      String libJarsCsv = copyFromLocal(fs, appPath, localJarFiles.toArray(new String[] {}));

      LOG.info("libjars: {}", libJarsCsv);
      dag.getAttributes().put(LogicalPlan.LIBRARY_JARS, libJarsCsv);
      LaunchContainerRunnable.addFilesToLocalResources(
          LocalResourceType.FILE, libJarsCsv, localResources, fs);

      if (archives != null) {
        String[] localFiles = archives.split(",");
        String archivesCsv = copyFromLocal(fs, appPath, localFiles);
        LOG.info("archives: {}", archivesCsv);
        dag.getAttributes().put(LogicalPlan.ARCHIVES, archivesCsv);
        LaunchContainerRunnable.addFilesToLocalResources(
            LocalResourceType.ARCHIVE, archivesCsv, localResources, fs);
      }

      if (files != null) {
        String[] localFiles = files.split(",");
        String filesCsv = copyFromLocal(fs, appPath, localFiles);
        LOG.info("files: {}", filesCsv);
        dag.getAttributes().put(LogicalPlan.FILES, filesCsv);
        LaunchContainerRunnable.addFilesToLocalResources(
            LocalResourceType.FILE, filesCsv, localResources, fs);
      }

      dag.getAttributes().put(LogicalPlan.APPLICATION_PATH, appPath.toString());
      if (dag.getAttributes().get(OperatorContext.STORAGE_AGENT) == null) {
          /* which would be the most likely case */
        Path checkpointPath = new Path(appPath, LogicalPlan.SUBDIR_CHECKPOINTS);
        // use conf client side to pickup any proxy settings from dt-site.xml
        dag.setAttribute(
            OperatorContext.STORAGE_AGENT, new FSStorageAgent(checkpointPath.toString(), conf));
      }
      if (dag.getAttributes().get(LogicalPlan.CONTAINER_OPTS_CONFIGURATOR) == null) {
        dag.setAttribute(
            LogicalPlan.CONTAINER_OPTS_CONFIGURATOR, new BasicContainerOptConfigurator());
      }

      // Set the log4j properties if needed
      if (!log4jPropFile.isEmpty()) {
        Path log4jSrc = new Path(log4jPropFile);
        Path log4jDst = new Path(appPath, "log4j.props");
        fs.copyFromLocalFile(false, true, log4jSrc, log4jDst);
        FileStatus log4jFileStatus = fs.getFileStatus(log4jDst);
        LocalResource log4jRsrc = Records.newRecord(LocalResource.class);
        log4jRsrc.setType(LocalResourceType.FILE);
        log4jRsrc.setVisibility(LocalResourceVisibility.APPLICATION);
        log4jRsrc.setResource(ConverterUtils.getYarnUrlFromURI(log4jDst.toUri()));
        log4jRsrc.setTimestamp(log4jFileStatus.getModificationTime());
        log4jRsrc.setSize(log4jFileStatus.getLen());
        localResources.put("log4j.properties", log4jRsrc);
      }

      if (originalAppId != null) {
        Path origAppPath = new Path(appsBasePath, this.originalAppId);
        LOG.info("Restart from {}", origAppPath);
        copyInitialState(origAppPath);
      }

      // push logical plan to DFS location
      Path cfgDst = new Path(appPath, LogicalPlan.SER_FILE_NAME);
      FSDataOutputStream outStream = fs.create(cfgDst, true);
      LogicalPlan.write(this.dag, outStream);
      outStream.close();

      Path launchConfigDst = new Path(appPath, LogicalPlan.LAUNCH_CONFIG_FILE_NAME);
      outStream = fs.create(launchConfigDst, true);
      conf.writeXml(outStream);
      outStream.close();

      FileStatus topologyFileStatus = fs.getFileStatus(cfgDst);
      LocalResource topologyRsrc = Records.newRecord(LocalResource.class);
      topologyRsrc.setType(LocalResourceType.FILE);
      topologyRsrc.setVisibility(LocalResourceVisibility.APPLICATION);
      topologyRsrc.setResource(ConverterUtils.getYarnUrlFromURI(cfgDst.toUri()));
      topologyRsrc.setTimestamp(topologyFileStatus.getModificationTime());
      topologyRsrc.setSize(topologyFileStatus.getLen());
      localResources.put(LogicalPlan.SER_FILE_NAME, topologyRsrc);

      // Set local resource info into app master container launch context
      amContainer.setLocalResources(localResources);

      // Set the necessary security tokens as needed
      // amContainer.setContainerTokens(containerToken);
      // Set the env variables to be setup in the env where the application master will be run
      LOG.info("Set the environment for the application master");
      Map<String, String> env = new HashMap<String, String>();

      // Add application jar(s) location to classpath
      // At some point we should not be required to add
      // the hadoop specific classpaths to the env.
      // It should be provided out of the box.
      // For now setting all required classpaths including
      // the classpath to "." for the application jar(s)
      // including ${CLASSPATH} will duplicate the class path in app master, removing it for now
      // StringBuilder classPathEnv = new StringBuilder("${CLASSPATH}:./*");
      StringBuilder classPathEnv = new StringBuilder("./*");
      String classpath = conf.get(YarnConfiguration.YARN_APPLICATION_CLASSPATH);
      for (String c :
          StringUtils.isBlank(classpath)
              ? YarnConfiguration.DEFAULT_YARN_APPLICATION_CLASSPATH
              : classpath.split(",")) {
        if (c.equals("$HADOOP_CLIENT_CONF_DIR")) {
          // SPOI-2501
          continue;
        }
        classPathEnv.append(':');
        classPathEnv.append(c.trim());
      }
      env.put("CLASSPATH", classPathEnv.toString());
      // propagate to replace node managers user name (effective in non-secure mode)
      env.put("HADOOP_USER_NAME", UserGroupInformation.getLoginUser().getUserName());

      amContainer.setEnvironment(env);

      // Set the necessary command to execute the application master
      ArrayList<CharSequence> vargs = new ArrayList<CharSequence>(30);

      // Set java executable command
      LOG.info("Setting up app master command");
      vargs.add(javaCmd);
      if (dag.isDebug()) {
        vargs.add("-agentlib:jdwp=transport=dt_socket,server=y,suspend=n");
      }
      // Set Xmx based on am memory size
      // default heap size 75% of total memory
      vargs.add("-Xmx" + (amMemory * 3 / 4) + "m");
      vargs.add("-XX:+HeapDumpOnOutOfMemoryError");
      vargs.add("-XX:HeapDumpPath=/tmp/dt-heap-" + appId.getId() + ".bin");
      vargs.add("-Dhadoop.root.logger=" + (dag.isDebug() ? "DEBUG" : "INFO") + ",RFA");
      vargs.add("-Dhadoop.log.dir=" + ApplicationConstants.LOG_DIR_EXPANSION_VAR);
      vargs.add(String.format("-D%s=%s", StreamingContainer.PROP_APP_PATH, dag.assertAppPath()));
      if (dag.isDebug()) {
        vargs.add("-Dlog4j.debug=true");
      }

      String loggersLevel = conf.get(DTLoggerFactory.DT_LOGGERS_LEVEL);
      if (loggersLevel != null) {
        vargs.add(String.format("-D%s=%s", DTLoggerFactory.DT_LOGGERS_LEVEL, loggersLevel));
      }
      vargs.add(StreamingAppMaster.class.getName());
      vargs.add("1>" + ApplicationConstants.LOG_DIR_EXPANSION_VAR + "/AppMaster.stdout");
      vargs.add("2>" + ApplicationConstants.LOG_DIR_EXPANSION_VAR + "/AppMaster.stderr");

      // Get final command
      StringBuilder command = new StringBuilder(9 * vargs.size());
      for (CharSequence str : vargs) {
        command.append(str).append(" ");
      }

      LOG.info("Completed setting up app master command " + command.toString());
      List<String> commands = new ArrayList<String>();
      commands.add(command.toString());
      amContainer.setCommands(commands);

      // Set up resource type requirements
      // For now, only memory is supported so we set memory requirements
      Resource capability = Records.newRecord(Resource.class);
      capability.setMemory(amMemory);
      appContext.setResource(capability);

      // Service data is a binary blob that can be passed to the application
      // Not needed in this scenario
      // amContainer.setServiceData(serviceData);
      appContext.setAMContainerSpec(amContainer);

      // Set the priority for the application master
      Priority pri = Records.newRecord(Priority.class);
      pri.setPriority(amPriority);
      appContext.setPriority(pri);
      // Set the queue to which this application is to be submitted in the RM
      appContext.setQueue(queueName);

      // Submit the application to the applications manager
      // SubmitApplicationResponse submitResp = rmClient.submitApplication(appRequest);
      // Ignore the response as either a valid response object is returned on success
      // or an exception thrown to denote some form of a failure
      String specStr =
          Objects.toStringHelper("Submitting application: ")
              .add("name", appContext.getApplicationName())
              .add("queue", appContext.getQueue())
              .add("user", UserGroupInformation.getLoginUser())
              .add("resource", appContext.getResource())
              .toString();
      LOG.info(specStr);
      if (dag.isDebug()) {
        // LOG.info("Full submission context: " + appContext);
      }
      yarnClient.submitApplication(appContext);
    } finally {
      fs.close();
    }
  }
Beispiel #3
0
  public int run(String[] args) throws Exception {
    // printUsage();
    /*
     * SETUP
     */
    Configuration argConf = getConf();
    Hashtable<String, String> confArg = new Hashtable<String, String>();
    setup(confArg, argConf);
    Date currentTime = new Date();
    Date endDate = new Date(new Long(confArg.get("timestamp_stop")));
    Boolean full_run = confArg.get("intermediate").matches("(?i).*true.*");
    Boolean quick_add = confArg.get("quick_add").matches("(?i).*true.*");
    logger.info("Running GeStore");

    // ZooKeeper setup
    Configuration config = HBaseConfiguration.create();
    zkWatcher = new ZooKeeperWatcher(config, "Testing", new HBaseAdmin(config));
    zkInstance =
        new ZooKeeper(
            ZKConfig.getZKQuorumServersString(config),
            config.getInt("zookeeper.session.timeout", -1),
            zkWatcher);

    if (!confArg.get("task_id").isEmpty()) {
      confArg.put("temp_path", confArg.get("temp_path") + confArg.get("task_id"));
    }

    String lockRequest = confArg.get("file_id");
    if (!confArg.get("run_id").isEmpty())
      lockRequest = lockRequest + "_" + confArg.get("run_id") + "_";
    if (!confArg.get("task_id").isEmpty())
      lockRequest = lockRequest + "_" + confArg.get("task_id") + "_";

    // Get type of movement
    toFrom type_move = checkArgs(confArg);
    if (type_move == toFrom.LOCAL2REMOTE && !confArg.get("format").equals("unknown")) {
      List<String> arguments = new ArrayList<String>();
      arguments.add("-Dinput=" + confArg.get("local_path"));
      arguments.add("-Dtable=" + confArg.get("file_id"));
      arguments.add("-Dtimestamp=" + confArg.get("timestamp_stop"));
      arguments.add("-Dtype=" + confArg.get("format"));
      arguments.add("-Dtarget_dir=" + confArg.get("base_path") + "_" + confArg.get("file_id"));
      arguments.add("-Dtemp_hdfs_path=" + confArg.get("temp_path"));
      arguments.add("-Drun_id=" + confArg.get("run_id"));
      if (!confArg.get("run_id").isEmpty()) arguments.add("-Drun_id=" + confArg.get("run_id"));
      if (!confArg.get("task_id").isEmpty()) arguments.add("-Dtask_id=" + confArg.get("task_id"));
      if (quick_add) arguments.add("-Dquick_add=" + confArg.get("quick_add"));
      String lockName = lock(lockRequest);
      String[] argumentString = arguments.toArray(new String[arguments.size()]);
      adddb.main(argumentString);
      unlock(lockName);
      System.exit(0);
    }

    // Database registration

    dbutil db_util = new dbutil(config);
    db_util.register_database(confArg.get("db_name_files"), true);
    db_util.register_database(confArg.get("db_name_runs"), true);
    db_util.register_database(confArg.get("db_name_updates"), true);
    FileSystem hdfs = FileSystem.get(config);
    FileSystem localFS = FileSystem.getLocal(config);

    // Get source type
    confArg.put("source", getSource(db_util, confArg.get("db_name_files"), confArg.get("file_id")));
    confArg.put(
        "database", isDatabase(db_util, confArg.get("db_name_files"), confArg.get("file_id")));
    if (!confArg.get("source").equals("local")
        && type_move == toFrom.REMOTE2LOCAL
        && !confArg.get("timestamp_stop").equals(Integer.toString(Integer.MAX_VALUE))) {
      confArg.put("timestamp_stop", Long.toString(latestVersion(confArg, db_util)));
    }

    /*
     * Get previous timestamp
     */
    Get run_id_get = new Get(confArg.get("run_id").getBytes());
    Result run_get = db_util.doGet(confArg.get("db_name_runs"), run_id_get);
    KeyValue run_file_prev =
        run_get.getColumnLatest(
            "d".getBytes(), (confArg.get("file_id") + "_db_timestamp").getBytes());
    String last_timestamp = new String("0");
    if (null != run_file_prev && !confArg.get("source").equals("local")) {
      long last_timestamp_real = run_file_prev.getTimestamp();
      Long current_timestamp = new Long(confArg.get("timestamp_real"));
      if ((current_timestamp - last_timestamp_real) > 36000) {
        last_timestamp = new String(run_file_prev.getValue());
        Integer lastTimestamp = new Integer(last_timestamp);
        lastTimestamp += 1;
        last_timestamp = lastTimestamp.toString();
        logger.info("Last timestamp: " + last_timestamp + " End data: " + endDate);
        Date last_run = new Date(run_file_prev.getTimestamp());
        if (last_run.before(endDate) && !full_run) {
          confArg.put("timestamp_start", last_timestamp);
        }
      }
    }

    Integer tse = new Integer(confArg.get("timestamp_stop"));
    Integer tss = new Integer(confArg.get("timestamp_start"));
    if (tss > tse) {
      logger.info("No new version of requested file.");
      return 0;
    }

    /*
     * Generate file
     */

    String lockName = lock(lockRequest);

    Get file_id_get = new Get(confArg.get("file_id").getBytes());
    Result file_get = db_util.doGet(confArg.get("db_name_files"), file_id_get);
    if (!file_get.isEmpty()) {
      boolean found =
          hasFile(
              db_util,
              hdfs,
              confArg.get("db_name_files"),
              confArg.get("file_id"),
              getFullPath(confArg));
      if (confArg.get("source").equals("fullfile")) {
        found = false;
      }
      String filenames_put =
          getFileNames(
              db_util, confArg.get("db_name_files"), confArg.get("file_id"), getFullPath(confArg));
      // Filename not found in file database
      if (!found && type_move == toFrom.REMOTE2LOCAL) {
        if (!confArg.get("source").equals("local")) {
          // Generate intermediate file
          if (getFile(hdfs, confArg, db_util) == null) {
            unlock(lockName);
            return 1;
          }
          // Put generated file into file database
          if (!confArg.get("format").equals("fullfile")) {
            putFileEntry(
                db_util,
                hdfs,
                confArg.get("db_name_files"),
                confArg.get("file_id"),
                confArg.get("full_file_name"),
                confArg.get("source"));
          }
        } else {
          logger.warn("Remote file not found, and cannot be generated! File: " + confArg);
          unlock(lockName);
          return 1;
        }
      }
    } else {
      if (type_move == toFrom.REMOTE2LOCAL) {
        logger.warn("Remote file not found, and cannot be generated.");
        unlock(lockName);
        return 1;
      }
    }

    /*
     * Copy file
     * Update tables
     */

    if (type_move == toFrom.LOCAL2REMOTE) {
      if (!confArg.get("format").equals("fullfile")) {
        putFileEntry(
            db_util,
            hdfs,
            confArg.get("db_name_files"),
            confArg.get("file_id"),
            getFullPath(confArg),
            confArg.get("source"));
      }
      putRunEntry(
          db_util,
          confArg.get("db_name_runs"),
          confArg.get("run_id"),
          confArg.get("file_id"),
          confArg.get("type"),
          confArg.get("timestamp_real"),
          confArg.get("timestamp_stop"),
          getFullPath(confArg),
          confArg.get("delimiter"));
      hdfs.copyFromLocalFile(new Path(confArg.get("local_path")), new Path(getFullPath(confArg)));
    } else if (type_move == toFrom.REMOTE2LOCAL) {
      FileStatus[] files = hdfs.globStatus(new Path(getFullPath(confArg) + "*"));
      putRunEntry(
          db_util,
          confArg.get("db_name_runs"),
          confArg.get("run_id"),
          confArg.get("file_id"),
          confArg.get("type"),
          confArg.get("timestamp_real"),
          confArg.get("timestamp_stop"),
          getFullPath(confArg),
          confArg.get("delimiter"));
      unlock(lockName);
      for (FileStatus file : files) {
        Path cur_file = file.getPath();
        Path cur_local_path =
            new Path(new String(confArg.get("local_path") + confArg.get("file_id")));
        String suffix = getSuffix(getFileName(confArg), cur_file.getName());
        if (suffix.length() > 0) {
          cur_local_path = cur_local_path.suffix(new String("." + suffix));
        }
        if (confArg.get("copy").equals("true")) {
          String crc = hdfs.getFileChecksum(cur_file).toString();
          if (checksumLocalTest(cur_local_path, crc)) {
            continue;
          } else {
            hdfs.copyToLocalFile(cur_file, cur_local_path);
            writeChecksum(cur_local_path, crc);
          }
        } else {
          System.out.println(cur_local_path + "\t" + cur_file);
        }
      }
    }
    unlock(lockName);
    return 0;
  }