Examples with MoveWork org.apache.hadoop.hive.ql.plan.MoveWork used on opensource projects

Example 1 with MoveWork

use of org.apache.hadoop.hive.ql.plan.MoveWork in project hive by apache.

the class GenMapRedUtils method createMRWorkForMergingFiles.

/**
   * @param fsInput The FileSink operator.
   * @param ctx The MR processing context.
   * @param finalName the final destination path the merge job should output.
   * @param dependencyTask
   * @param mvTasks
   * @param conf
   * @param currTask
   * @throws SemanticException

   * create a Map-only merge job using CombineHiveInputFormat for all partitions with
   * following operators:
   *          MR job J0:
   *          ...
   *          |
   *          v
   *          FileSinkOperator_1 (fsInput)
   *          |
   *          v
   *          Merge job J1:
   *          |
   *          v
   *          TableScan (using CombineHiveInputFormat) (tsMerge)
   *          |
   *          v
   *          FileSinkOperator (fsMerge)
   *
   *          Here the pathToPartitionInfo & pathToAlias will remain the same, which means the paths
   *          do
   *          not contain the dynamic partitions (their parent). So after the dynamic partitions are
   *          created (after the first job finished before the moveTask or ConditionalTask start),
   *          we need to change the pathToPartitionInfo & pathToAlias to include the dynamic
   *          partition
   *          directories.
   *
   */
public static void createMRWorkForMergingFiles(FileSinkOperator fsInput, Path finalName, DependencyCollectionTask dependencyTask, List<Task<MoveWork>> mvTasks, HiveConf conf, Task<? extends Serializable> currTask) throws SemanticException {
    //
    // 1. create the operator tree
    //
    FileSinkDesc fsInputDesc = fsInput.getConf();
    // Create a TableScan operator
    RowSchema inputRS = fsInput.getSchema();
    TableScanOperator tsMerge = GenMapRedUtils.createTemporaryTableScanOperator(fsInput.getCompilationOpContext(), inputRS);
    // Create a FileSink operator
    TableDesc ts = (TableDesc) fsInputDesc.getTableInfo().clone();
    FileSinkDesc fsOutputDesc = new FileSinkDesc(finalName, ts, conf.getBoolVar(ConfVars.COMPRESSRESULT));
    FileSinkOperator fsOutput = (FileSinkOperator) OperatorFactory.getAndMakeChild(fsOutputDesc, inputRS, tsMerge);
    // If the input FileSinkOperator is a dynamic partition enabled, the tsMerge input schema
    // needs to include the partition column, and the fsOutput should have
    // a DynamicPartitionCtx to indicate that it needs to dynamically partitioned.
    DynamicPartitionCtx dpCtx = fsInputDesc.getDynPartCtx();
    if (dpCtx != null && dpCtx.getNumDPCols() > 0) {
        // adding DP ColumnInfo to the RowSchema signature
        ArrayList<ColumnInfo> signature = inputRS.getSignature();
        String tblAlias = fsInputDesc.getTableInfo().getTableName();
        for (String dpCol : dpCtx.getDPColNames()) {
            ColumnInfo colInfo = new ColumnInfo(dpCol, // all partition column type should be string
            TypeInfoFactory.stringTypeInfo, tblAlias, // partition column is virtual column
            true);
            signature.add(colInfo);
        }
        inputRS.setSignature(signature);
        // create another DynamicPartitionCtx, which has a different input-to-DP column mapping
        DynamicPartitionCtx dpCtx2 = new DynamicPartitionCtx(dpCtx);
        fsOutputDesc.setDynPartCtx(dpCtx2);
        // update the FileSinkOperator to include partition columns
        usePartitionColumns(fsInputDesc.getTableInfo().getProperties(), dpCtx.getDPColNames());
    } else {
        // non-partitioned table
        fsInputDesc.getTableInfo().getProperties().remove(org.apache.hadoop.hive.metastore.api.hive_metastoreConstants.META_TABLE_PARTITION_COLUMNS);
    }
    //
    // 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) && fsInputDesc.getTableInfo().getInputFileFormatClass().equals(RCFileInputFormat.class)) || (conf.getBoolVar(ConfVars.HIVEMERGEORCFILESTRIPELEVEL) && fsInputDesc.getTableInfo().getInputFileFormatClass().equals(OrcInputFormat.class))) {
        cplan = GenMapRedUtils.createMergeTask(fsInputDesc, finalName, dpCtx != null && dpCtx.getNumDPCols() > 0, fsInput.getCompilationOpContext());
        if (conf.getVar(ConfVars.HIVE_EXECUTION_ENGINE).equals("tez")) {
            work = new TezWork(conf.getVar(HiveConf.ConfVars.HIVEQUERYID), conf);
            cplan.setName("File Merge");
            ((TezWork) work).add(cplan);
        } else if (conf.getVar(ConfVars.HIVE_EXECUTION_ENGINE).equals("spark")) {
            work = new SparkWork(conf.getVar(HiveConf.ConfVars.HIVEQUERYID));
            cplan.setName("Spark Merge File Work");
            ((SparkWork) work).add(cplan);
        } else {
            work = cplan;
        }
    } else {
        cplan = createMRWorkForMergingFiles(conf, tsMerge, fsInputDesc);
        if (conf.getVar(ConfVars.HIVE_EXECUTION_ENGINE).equals("tez")) {
            work = new TezWork(conf.getVar(HiveConf.ConfVars.HIVEQUERYID), conf);
            cplan.setName("File Merge");
            ((TezWork) work).add(cplan);
        } else if (conf.getVar(ConfVars.HIVE_EXECUTION_ENGINE).equals("spark")) {
            work = new SparkWork(conf.getVar(HiveConf.ConfVars.HIVEQUERYID));
            cplan.setName("Spark Merge File Work");
            ((SparkWork) work).add(cplan);
        } else {
            work = new MapredWork();
            ((MapredWork) work).setMapWork(cplan);
        }
    }
    // use CombineHiveInputFormat for map-only merging
    cplan.setInputformat("org.apache.hadoop.hive.ql.io.CombineHiveInputFormat");
    // NOTE: we should gather stats in MR1 rather than MR2 at merge job since we don't
    // know if merge MR2 will be triggered at execution time
    Task<MoveWork> mvTask = GenMapRedUtils.findMoveTask(mvTasks, fsOutput);
    ConditionalTask cndTsk = GenMapRedUtils.createCondTask(conf, currTask, dummyMv, work, fsInputDesc.getFinalDirName(), finalName, mvTask, dependencyTask);
    // keep the dynamic partition context in conditional task resolver context
    ConditionalResolverMergeFilesCtx mrCtx = (ConditionalResolverMergeFilesCtx) cndTsk.getResolverCtx();
    mrCtx.setDPCtx(fsInputDesc.getDynPartCtx());
    mrCtx.setLbCtx(fsInputDesc.getLbCtx());
}

Example 2 with MoveWork

use of org.apache.hadoop.hive.ql.plan.MoveWork in project hive by apache.

the class ImportSemanticAnalyzer method addSinglePartition.

private static Task<?> addSinglePartition(URI fromURI, FileSystem fs, ImportTableDesc tblDesc, Table table, Warehouse wh, AddPartitionDesc addPartitionDesc, ReplicationSpec replicationSpec, EximUtil.SemanticAnalyzerWrapperContext x) throws MetaException, IOException, HiveException {
    AddPartitionDesc.OnePartitionDesc partSpec = addPartitionDesc.getPartition(0);
    if (tblDesc.isExternal() && tblDesc.getLocation() == null) {
        x.getLOG().debug("Importing in-place: adding AddPart for partition " + partSpecToString(partSpec.getPartSpec()));
        // addPartitionDesc already has the right partition location
        Task<?> addPartTask = TaskFactory.get(new DDLWork(x.getInputs(), x.getOutputs(), addPartitionDesc), x.getConf());
        return addPartTask;
    } else {
        String srcLocation = partSpec.getLocation();
        fixLocationInPartSpec(fs, tblDesc, table, wh, replicationSpec, partSpec, x);
        x.getLOG().debug("adding dependent CopyWork/AddPart/MoveWork for partition " + partSpecToString(partSpec.getPartSpec()) + " with source location: " + srcLocation);
        Path tgtLocation = new Path(partSpec.getLocation());
        Path tmpPath = x.getCtx().getExternalTmpPath(tgtLocation);
        Task<?> copyTask = ReplCopyTask.getLoadCopyTask(replicationSpec, new Path(srcLocation), tmpPath, x.getConf());
        Task<?> addPartTask = TaskFactory.get(new DDLWork(x.getInputs(), x.getOutputs(), addPartitionDesc), x.getConf());
        LoadTableDesc loadTableWork = new LoadTableDesc(tmpPath, Utilities.getTableDesc(table), partSpec.getPartSpec(), true);
        loadTableWork.setInheritTableSpecs(false);
        Task<?> loadPartTask = TaskFactory.get(new MoveWork(x.getInputs(), x.getOutputs(), loadTableWork, null, false), x.getConf());
        copyTask.addDependentTask(loadPartTask);
        addPartTask.addDependentTask(loadPartTask);
        x.getTasks().add(copyTask);
        return addPartTask;
    }
}

Example 3 with MoveWork

use of org.apache.hadoop.hive.ql.plan.MoveWork in project hive by apache.

the class ImportSemanticAnalyzer method loadTable.

private static Task<?> loadTable(URI fromURI, Table table, boolean replace, Path tgtPath, ReplicationSpec replicationSpec, EximUtil.SemanticAnalyzerWrapperContext x) {
    Path dataPath = new Path(fromURI.toString(), EximUtil.DATA_PATH_NAME);
    Path tmpPath = x.getCtx().getExternalTmpPath(tgtPath);
    Task<?> copyTask = ReplCopyTask.getLoadCopyTask(replicationSpec, dataPath, tmpPath, x.getConf());
    LoadTableDesc loadTableWork = new LoadTableDesc(tmpPath, Utilities.getTableDesc(table), new TreeMap<String, String>(), replace);
    Task<?> loadTableTask = TaskFactory.get(new MoveWork(x.getInputs(), x.getOutputs(), loadTableWork, null, false), x.getConf());
    copyTask.addDependentTask(loadTableTask);
    x.getTasks().add(copyTask);
    return loadTableTask;
}

Example 4 with MoveWork

use of org.apache.hadoop.hive.ql.plan.MoveWork in project hive by apache.

the class DDLTask method generateAddMmTasks.

private List<Task<?>> generateAddMmTasks(Table tbl) throws HiveException {
    // We will move all the files in the table/partition directories into the first MM
    // directory, then commit the first write ID.
    List<Path> srcs = new ArrayList<>(), tgts = new ArrayList<>();
    long mmWriteId = 0;
    try {
        HiveTxnManager txnManager = SessionState.get().getTxnMgr();
        if (txnManager.isTxnOpen()) {
            mmWriteId = txnManager.getTableWriteId(tbl.getDbName(), tbl.getTableName());
        } else {
            txnManager.openTxn(new Context(conf), conf.getUser());
            mmWriteId = txnManager.getTableWriteId(tbl.getDbName(), tbl.getTableName());
            txnManager.commitTxn();
        }
    } catch (Exception e) {
        String errorMessage = "FAILED: Error in acquiring locks: " + e.getMessage();
        console.printError(errorMessage, "\n" + org.apache.hadoop.util.StringUtils.stringifyException(e));
    }
    int stmtId = 0;
    String mmDir = AcidUtils.deltaSubdir(mmWriteId, mmWriteId, stmtId);
    Hive db = getHive();
    if (tbl.getPartitionKeys().size() > 0) {
        PartitionIterable parts = new PartitionIterable(db, tbl, null, HiveConf.getIntVar(conf, ConfVars.METASTORE_BATCH_RETRIEVE_MAX));
        Iterator<Partition> partIter = parts.iterator();
        while (partIter.hasNext()) {
            Partition part = partIter.next();
            checkMmLb(part);
            Path src = part.getDataLocation(), tgt = new Path(src, mmDir);
            srcs.add(src);
            tgts.add(tgt);
            if (Utilities.FILE_OP_LOGGER.isTraceEnabled()) {
                Utilities.FILE_OP_LOGGER.trace("Will move " + src + " to " + tgt);
            }
        }
    } else {
        checkMmLb(tbl);
        Path src = tbl.getDataLocation(), tgt = new Path(src, mmDir);
        srcs.add(src);
        tgts.add(tgt);
        if (Utilities.FILE_OP_LOGGER.isTraceEnabled()) {
            Utilities.FILE_OP_LOGGER.trace("Will move " + src + " to " + tgt);
        }
    }
    // Don't set inputs and outputs - the locks have already been taken so it's pointless.
    MoveWork mw = new MoveWork(null, null, null, null, false);
    mw.setMultiFilesDesc(new LoadMultiFilesDesc(srcs, tgts, true, null, null));
    ImportCommitWork icw = new ImportCommitWork(tbl.getDbName(), tbl.getTableName(), mmWriteId, stmtId);
    Task<?> mv = TaskFactory.get(mw), ic = TaskFactory.get(icw);
    mv.addDependentTask(ic);
    return Lists.<Task<?>>newArrayList(mv);
}

Example 5 with MoveWork

use of org.apache.hadoop.hive.ql.plan.MoveWork in project hive by apache.

the class DDLSemanticAnalyzer method analyzeAlterTablePartMergeFiles.

private void analyzeAlterTablePartMergeFiles(ASTNode ast, String tableName, HashMap<String, String> partSpec) throws SemanticException {
    AlterTablePartMergeFilesDesc mergeDesc = new AlterTablePartMergeFilesDesc(tableName, partSpec);
    List<Path> inputDir = new ArrayList<Path>();
    Path oldTblPartLoc = null;
    Path newTblPartLoc = null;
    Table tblObj = null;
    ListBucketingCtx lbCtx = null;
    try {
        tblObj = getTable(tableName);
        // TODO: we should probably block all ACID tables here.
        if (AcidUtils.isInsertOnlyTable(tblObj.getParameters())) {
            throw new SemanticException("Merge is not supported for MM tables");
        }
        mergeDesc.setTableDesc(Utilities.getTableDesc(tblObj));
        List<String> bucketCols = null;
        Class<? extends InputFormat> inputFormatClass = null;
        boolean isArchived = false;
        if (tblObj.isPartitioned()) {
            if (partSpec == null) {
                throw new SemanticException("source table " + tableName + " is partitioned but no partition desc found.");
            } else {
                Partition part = getPartition(tblObj, partSpec, false);
                if (part == null) {
                    throw new SemanticException("source table " + tableName + " is partitioned but partition not found.");
                }
                bucketCols = part.getBucketCols();
                inputFormatClass = part.getInputFormatClass();
                isArchived = ArchiveUtils.isArchived(part);
                Path tabPath = tblObj.getPath();
                Path partPath = part.getDataLocation();
                // if the table is in a different dfs than the partition,
                // replace the partition's dfs with the table's dfs.
                newTblPartLoc = new Path(tabPath.toUri().getScheme(), tabPath.toUri().getAuthority(), partPath.toUri().getPath());
                oldTblPartLoc = partPath;
                lbCtx = constructListBucketingCtx(part.getSkewedColNames(), part.getSkewedColValues(), part.getSkewedColValueLocationMaps(), part.isStoredAsSubDirectories(), conf);
            }
        } else {
            inputFormatClass = tblObj.getInputFormatClass();
            bucketCols = tblObj.getBucketCols();
            // input and output are the same
            oldTblPartLoc = tblObj.getPath();
            newTblPartLoc = tblObj.getPath();
            lbCtx = constructListBucketingCtx(tblObj.getSkewedColNames(), tblObj.getSkewedColValues(), tblObj.getSkewedColValueLocationMaps(), tblObj.isStoredAsSubDirectories(), conf);
        }
        // throw a HiveException for other than rcfile and orcfile.
        if (!((inputFormatClass.equals(RCFileInputFormat.class) || (inputFormatClass.equals(OrcInputFormat.class))))) {
            throw new SemanticException(ErrorMsg.CONCATENATE_UNSUPPORTED_FILE_FORMAT.getMsg());
        }
        mergeDesc.setInputFormatClass(inputFormatClass);
        // throw a HiveException if the table/partition is bucketized
        if (bucketCols != null && bucketCols.size() > 0) {
            throw new SemanticException(ErrorMsg.CONCATENATE_UNSUPPORTED_TABLE_BUCKETED.getMsg());
        }
        // throw a HiveException if the table/partition is archived
        if (isArchived) {
            throw new SemanticException(ErrorMsg.CONCATENATE_UNSUPPORTED_PARTITION_ARCHIVED.getMsg());
        }
        // violating which can cause data loss
        if (tblObj.isNonNative()) {
            throw new SemanticException(ErrorMsg.CONCATENATE_UNSUPPORTED_TABLE_NON_NATIVE.getMsg());
        }
        if (tblObj.getTableType() != TableType.MANAGED_TABLE) {
            throw new SemanticException(ErrorMsg.CONCATENATE_UNSUPPORTED_TABLE_NOT_MANAGED.getMsg());
        }
        // transactional tables are compacted and no longer needs to be bucketed, so not safe for merge/concatenation
        boolean isAcid = AcidUtils.isTransactionalTable(tblObj);
        if (isAcid) {
            throw new SemanticException(ErrorMsg.CONCATENATE_UNSUPPORTED_TABLE_TRANSACTIONAL.getMsg());
        }
        inputDir.add(oldTblPartLoc);
        mergeDesc.setInputDir(inputDir);
        mergeDesc.setLbCtx(lbCtx);
        addInputsOutputsAlterTable(tableName, partSpec, AlterTableTypes.MERGEFILES);
        DDLWork ddlWork = new DDLWork(getInputs(), getOutputs(), mergeDesc);
        ddlWork.setNeedLock(true);
        Task<? extends Serializable> mergeTask = TaskFactory.get(ddlWork);
        TableDesc tblDesc = Utilities.getTableDesc(tblObj);
        Path queryTmpdir = ctx.getExternalTmpPath(newTblPartLoc);
        mergeDesc.setOutputDir(queryTmpdir);
        // No need to handle MM tables - unsupported path.
        LoadTableDesc ltd = new LoadTableDesc(queryTmpdir, tblDesc, partSpec == null ? new HashMap<>() : partSpec);
        ltd.setLbCtx(lbCtx);
        Task<MoveWork> moveTsk = TaskFactory.get(new MoveWork(null, null, ltd, null, false));
        mergeTask.addDependentTask(moveTsk);
        if (conf.getBoolVar(HiveConf.ConfVars.HIVESTATSAUTOGATHER)) {
            BasicStatsWork basicStatsWork;
            if (oldTblPartLoc.equals(newTblPartLoc)) {
                // If we're merging to the same location, we can avoid some metastore calls
                TableSpec tableSpec = new TableSpec(db, tableName, partSpec);
                basicStatsWork = new BasicStatsWork(tableSpec);
            } else {
                basicStatsWork = new BasicStatsWork(ltd);
            }
            basicStatsWork.setNoStatsAggregator(true);
            basicStatsWork.setClearAggregatorStats(true);
            StatsWork columnStatsWork = new StatsWork(tblObj, basicStatsWork, conf);
            Task<? extends Serializable> statTask = TaskFactory.get(columnStatsWork);
            moveTsk.addDependentTask(statTask);
        }
        rootTasks.add(mergeTask);
    } catch (Exception e) {
        throw new SemanticException(e);
    }
}