Examples of MoveWork


Examples of org.apache.hadoop.hive.ql.plan.MoveWork

      String queryTmpdir = ctx.getExternalTmpFileURI(newTblPartLoc.toUri());
      mergeDesc.setOutputDir(queryTmpdir);
      LoadTableDesc ltd = new LoadTableDesc(queryTmpdir, queryTmpdir, tblDesc,
          partSpec == null ? new HashMap<String, String>() : partSpec);
      ltd.setLbCtx(lbCtx);
      Task<MoveWork> moveTsk = TaskFactory.get(new MoveWork(null, null, ltd, null, false),
          conf);
      mergeTask.addDependentTask(moveTsk);

      if (conf.getBoolVar(HiveConf.ConfVars.HIVESTATSAUTOGATHER)) {
        StatsWork statDesc;
View Full Code Here

Examples of org.apache.hadoop.hive.ql.plan.MoveWork

   *          HiveConf
   */
  private void addStatsTask(FileSinkOperator nd, MoveTask mvTask,
      Task<? extends Serializable> currTask, HiveConf hconf) {

    MoveWork mvWork = ((MoveTask) mvTask).getWork();
    StatsWork statsWork = null;
    if (mvWork.getLoadTableWork() != null) {
      statsWork = new StatsWork(mvWork.getLoadTableWork());
    } else if (mvWork.getLoadFileWork() != null) {
      statsWork = new StatsWork(mvWork.getLoadFileWork());
    }
    assert statsWork != null : "Error when genereting StatsTask";
    statsWork.setStatsReliable(hconf.getBoolVar(ConfVars.HIVE_STATS_RELIABLE));
    MapredWork mrWork = (MapredWork) currTask.getWork();

View Full Code Here

Examples of org.apache.hadoop.hive.ql.plan.MoveWork

    }

    //
    // 2. Constructing a conditional task consisting of a move task and a map reduce task
    //
    MoveWork dummyMv = new MoveWork(null, null, null,
        new LoadFileDesc(fsInputDesc.getFinalDirName(), finalName, true, null, null), false);
    MapWork cplan;
    Serializable work;

    if (conf.getBoolVar(ConfVars.HIVEMERGERCFILEBLOCKLEVEL) &&
View Full Code Here

Examples of org.apache.hadoop.hive.ql.plan.MoveWork

  private Task<MoveWork> findMoveTask(
      List<Task<MoveWork>> mvTasks, FileSinkOperator fsOp) {
    // find the move task
    for (Task<MoveWork> mvTsk : mvTasks) {
      MoveWork mvWork = mvTsk.getWork();
      String srcDir = null;
      if (mvWork.getLoadFileWork() != null) {
        srcDir = mvWork.getLoadFileWork().getSourceDir();
      } else if (mvWork.getLoadTableWork() != null) {
        srcDir = mvWork.getLoadTableWork().getSourceDir();
      }

      String fsOpDirName = fsOp.getConf().getFinalDirName();
      if ((srcDir != null)
          && (srcDir.equalsIgnoreCase(fsOpDirName))) {
View Full Code Here

Examples of org.apache.hadoop.hive.ql.plan.MoveWork

      TableDesc tblDesc = Utilities.getTableDesc(tblObj);
      String queryTmpdir = ctx.getExternalTmpFileURI(new URI(tblPartLoc));
      mergeDesc.setOutputDir(queryTmpdir);
      LoadTableDesc ltd = new LoadTableDesc(queryTmpdir, queryTmpdir, tblDesc,
          partSpec == null ? new HashMap<String, String>() : partSpec);
      Task<MoveWork> moveTsk = TaskFactory.get(new MoveWork(null, null, ltd, null, false),
          conf);
      mergeTask.addDependentTask(moveTsk);
      tableSpec tablepart = new tableSpec(this.db, conf, tablePartAST);
      StatsWork statDesc = new StatsWork(tablepart);
      statDesc.setNoStatsAggregator(true);
View Full Code Here

Examples of org.apache.hadoop.hive.ql.plan.MoveWork

        tmpURI, false), conf);
    LoadTableDesc loadTableWork = new LoadTableDesc(tmpURI.toString(),
        ctx.getExternalTmpFileURI(fromURI),
        Utilities.getTableDesc(table), new TreeMap<String, String>(),
        false);
    Task<?> loadTableTask = TaskFactory.get(new MoveWork(getInputs(),
        getOutputs(), loadTableWork, null, false), conf);
    copyTask.addDependentTask(loadTableTask);
    rootTasks.add(copyTask);
    return loadTableTask;
  }
View Full Code Here

Examples of org.apache.hadoop.hive.ql.plan.MoveWork

      LoadTableDesc loadTableWork = new LoadTableDesc(tmpURI,
          ctx.getExternalTmpFileURI(fromURI),
          Utilities.getTableDesc(table),
          addPartitionDesc.getPartSpec(), true);
      loadTableWork.setInheritTableSpecs(false);
      Task<?> loadPartTask = TaskFactory.get(new MoveWork(
          getInputs(), getOutputs(), loadTableWork, null, false),
          conf);
      copyTask.addDependentTask(loadPartTask);
      addPartTask.addDependentTask(loadPartTask);
      rootTasks.add(copyTask);
View Full Code Here

Examples of org.apache.hadoop.hive.ql.plan.MoveWork

        globalLimitCtx.disableOpt();
      }

    } else {
      for (LoadTableDesc ltd : loadTableWork) {
        Task<MoveWork> tsk = TaskFactory.get(new MoveWork(null, null, ltd, null, false),
            conf);
        mvTask.add(tsk);
      }

      boolean oneLoadFile = true;
      for (LoadFileDesc lfd : loadFileWork) {
        if (qb.isCTAS()) {
          assert (oneLoadFile); // should not have more than 1 load file for
          // CTAS
          // make the movetask's destination directory the table's destination.
          String location = qb.getTableDesc().getLocation();
          if (location == null) {
            // get the table's default location
            Table dumpTable;
            Path targetPath;
            try {
              dumpTable = db.newTable(qb.getTableDesc().getTableName());
              Warehouse wh = new Warehouse(conf);
              targetPath = wh.getTablePath(db.getDatabase(dumpTable.getDbName()), dumpTable
                  .getTableName());
            } catch (HiveException e) {
              throw new SemanticException(e);
            } catch (MetaException e) {
              throw new SemanticException(e);
            }

            location = targetPath.toString();
          }
          lfd.setTargetDir(location);

          oneLoadFile = false;
        }
        mvTask.add(TaskFactory.get(new MoveWork(null, null, null, lfd, false),
            conf));
      }
    }

    // generate map reduce plans
View Full Code Here

Examples of org.apache.hadoop.hive.ql.plan.MoveWork

    LoadTableDesc loadTableWork = new LoadTableDesc(fromURI.toString(),
        loadTmpPath, Utilities.getTableDesc(ts.tableHandle), partSpec, isOverWrite);

    if (rTask != null) {
      rTask.addDependentTask(TaskFactory.get(new MoveWork(getInputs(),
          getOutputs(), loadTableWork, null, true), conf));
    } else {
      rTask = TaskFactory.get(new MoveWork(getInputs(), getOutputs(),
          loadTableWork, null, true), conf);
    }

    rootTasks.add(rTask);
  }
View Full Code Here

Examples of org.apache.hadoop.hive.ql.plan.MoveWork

   * @param hconf HiveConf
   */
  private void addStatsTask(FileSinkOperator nd, MoveTask mvTask,
      Task<? extends Serializable> currTask, HiveConf hconf) {

    MoveWork mvWork = ((MoveTask)mvTask).getWork();
    StatsWork statsWork = new StatsWork(mvWork.getLoadTableWork());
    MapredWork mrWork = (MapredWork) currTask.getWork();

    // AggKey in StatsWork is used for stats aggregation while StatsAggPrefix
    // in FileSinkDesc is used for stats publishing. They should be consistent.
    statsWork.setAggKey(((FileSinkOperator)nd).getConf().getStatsAggPrefix());
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.