Search in sources :

Example 46 with TableName

use of org.apache.hadoop.hive.common.TableName in project hive by apache.

the class AbstractAlterTableRenameAnalyzer method analyzeCommand.

@Override
protected void analyzeCommand(TableName tableName, Map<String, String> partitionSpec, ASTNode command) throws SemanticException {
    TableName target = getQualifiedTableName((ASTNode) command.getChild(0));
    AlterTableRenameDesc desc = new AlterTableRenameDesc(tableName, null, isView(), target.getNotEmptyDbTable());
    Table table = getTable(tableName.getNotEmptyDbTable(), true);
    if (AcidUtils.isTransactionalTable(table)) {
        setAcidDdlDesc(desc);
    }
    addInputsOutputsAlterTable(tableName, null, desc, desc.getType(), false);
    rootTasks.add(TaskFactory.get(new DDLWork(getInputs(), getOutputs(), desc)));
}
Also used : TableName(org.apache.hadoop.hive.common.TableName) Table(org.apache.hadoop.hive.ql.metadata.Table) DDLWork(org.apache.hadoop.hive.ql.ddl.DDLWork)

Example 47 with TableName

use of org.apache.hadoop.hive.common.TableName in project hive by apache.

the class HiveAlterHandler method alterTable.

@Override
public void alterTable(RawStore msdb, Warehouse wh, String catName, String dbname, String name, Table newt, EnvironmentContext environmentContext, IHMSHandler handler, String writeIdList) throws InvalidOperationException, MetaException {
    catName = normalizeIdentifier(catName);
    name = name.toLowerCase();
    dbname = dbname.toLowerCase();
    final boolean cascade;
    final boolean replDataLocationChanged;
    final boolean isReplicated;
    if ((environmentContext != null) && environmentContext.isSetProperties()) {
        cascade = StatsSetupConst.TRUE.equals(environmentContext.getProperties().get(StatsSetupConst.CASCADE));
        replDataLocationChanged = ReplConst.TRUE.equals(environmentContext.getProperties().get(ReplConst.REPL_DATA_LOCATION_CHANGED));
    } else {
        cascade = false;
        replDataLocationChanged = false;
    }
    if (newt == null) {
        throw new InvalidOperationException("New table is null");
    }
    String newTblName = newt.getTableName().toLowerCase();
    String newDbName = newt.getDbName().toLowerCase();
    if (!MetaStoreUtils.validateName(newTblName, handler.getConf())) {
        throw new InvalidOperationException(newTblName + " is not a valid object name");
    }
    String validate = MetaStoreServerUtils.validateTblColumns(newt.getSd().getCols());
    if (validate != null) {
        throw new InvalidOperationException("Invalid column " + validate);
    }
    // Validate bucketedColumns in new table
    List<String> bucketColumns = MetaStoreServerUtils.validateBucketColumns(newt.getSd());
    if (CollectionUtils.isNotEmpty(bucketColumns)) {
        String errMsg = "Bucket columns - " + bucketColumns.toString() + " doesn't match with any table columns";
        LOG.error(errMsg);
        throw new InvalidOperationException(errMsg);
    }
    Path srcPath = null;
    FileSystem srcFs;
    Path destPath = null;
    FileSystem destFs = null;
    boolean success = false;
    boolean dataWasMoved = false;
    boolean isPartitionedTable = false;
    Database olddb = null;
    Table oldt = null;
    List<TransactionalMetaStoreEventListener> transactionalListeners = handler.getTransactionalListeners();
    List<MetaStoreEventListener> listeners = handler.getListeners();
    Map<String, String> txnAlterTableEventResponses = Collections.emptyMap();
    try {
        boolean rename = false;
        List<Partition> parts;
        // Switching tables between catalogs is not allowed.
        if (!catName.equalsIgnoreCase(newt.getCatName())) {
            throw new InvalidOperationException("Tables cannot be moved between catalogs, old catalog" + catName + ", new catalog " + newt.getCatName());
        }
        // check if table with the new name already exists
        if (!newTblName.equals(name) || !newDbName.equals(dbname)) {
            if (msdb.getTable(catName, newDbName, newTblName, null) != null) {
                throw new InvalidOperationException("new table " + newDbName + "." + newTblName + " already exists");
            }
            rename = true;
        }
        msdb.openTransaction();
        // get old table
        // Note: we don't verify stats here; it's done below in alterTableUpdateTableColumnStats.
        olddb = msdb.getDatabase(catName, dbname);
        oldt = msdb.getTable(catName, dbname, name, null);
        if (oldt == null) {
            throw new InvalidOperationException("table " + TableName.getQualified(catName, dbname, name) + " doesn't exist");
        }
        validateTableChangesOnReplSource(olddb, oldt, newt, environmentContext);
        // On a replica this alter table will be executed only if old and new both the databases are
        // available and being replicated into. Otherwise, it will be either create or drop of table.
        isReplicated = HMSHandler.isDbReplicationTarget(olddb);
        if (oldt.getPartitionKeysSize() != 0) {
            isPartitionedTable = true;
        }
        // Throws InvalidOperationException if the new column types are not
        // compatible with the current column types.
        DefaultIncompatibleTableChangeHandler.get().allowChange(handler.getConf(), oldt, newt);
        // check that partition keys have not changed, except for virtual views
        // however, allow the partition comments to change
        boolean partKeysPartiallyEqual = checkPartialPartKeysEqual(oldt.getPartitionKeys(), newt.getPartitionKeys());
        if (!oldt.getTableType().equals(TableType.VIRTUAL_VIEW.toString())) {
            Map<String, String> properties = environmentContext.getProperties();
            if (properties == null || (properties != null && !Boolean.parseBoolean(properties.getOrDefault(HiveMetaHook.ALLOW_PARTITION_KEY_CHANGE, "false")))) {
                if (!partKeysPartiallyEqual) {
                    throw new InvalidOperationException("partition keys can not be changed.");
                }
            }
        }
        // 4) the table was not initially created with a specified location
        if (replDataLocationChanged || (rename && !oldt.getTableType().equals(TableType.VIRTUAL_VIEW.toString()) && (oldt.getSd().getLocation().compareTo(newt.getSd().getLocation()) == 0 || StringUtils.isEmpty(newt.getSd().getLocation())) && !MetaStoreUtils.isExternalTable(oldt))) {
            srcPath = new Path(oldt.getSd().getLocation());
            if (replDataLocationChanged) {
                // If data location is changed in replication flow, then new path was already set in
                // the newt. Also, it is as good as the data is moved and set dataWasMoved=true so that
                // location in partitions are also updated accordingly.
                // No need to validate if the destPath exists as in replication flow, data gets replicated
                // separately.
                destPath = new Path(newt.getSd().getLocation());
                dataWasMoved = true;
            } else {
                // Rename flow.
                // If a table was created in a user specified location using the DDL like
                // create table tbl ... location ...., it should be treated like an external table
                // in the table rename, its data location should not be changed. We can check
                // if the table directory was created directly under its database directory to tell
                // if it is such a table
                String oldtRelativePath = wh.getDatabaseManagedPath(olddb).toUri().relativize(srcPath.toUri()).toString();
                boolean tableInSpecifiedLoc = !oldtRelativePath.equalsIgnoreCase(name) && !oldtRelativePath.equalsIgnoreCase(name + Path.SEPARATOR);
                if (!tableInSpecifiedLoc) {
                    srcFs = wh.getFs(srcPath);
                    // get new location
                    Database db = msdb.getDatabase(catName, newDbName);
                    assert (isReplicated == HMSHandler.isDbReplicationTarget(db));
                    Path databasePath = constructRenamedPath(wh.getDatabaseManagedPath(db), srcPath);
                    destPath = new Path(databasePath, newTblName);
                    destFs = wh.getFs(destPath);
                    newt.getSd().setLocation(destPath.toString());
                    // check that src and dest are on the same file system
                    if (!FileUtils.equalsFileSystem(srcFs, destFs)) {
                        throw new InvalidOperationException("table new location " + destPath + " is on a different file system than the old location " + srcPath + ". This operation is not supported");
                    }
                    try {
                        if (destFs.exists(destPath)) {
                            throw new InvalidOperationException("New location for this table " + TableName.getQualified(catName, newDbName, newTblName) + " already exists : " + destPath);
                        }
                        // check that src exists and also checks permissions necessary, rename src to dest
                        if (srcFs.exists(srcPath) && wh.renameDir(srcPath, destPath, ReplChangeManager.shouldEnableCm(olddb, oldt))) {
                            dataWasMoved = true;
                        }
                    } catch (IOException | MetaException e) {
                        LOG.error("Alter Table operation for " + dbname + "." + name + " failed.", e);
                        throw new InvalidOperationException("Alter Table operation for " + dbname + "." + name + " failed to move data due to: '" + getSimpleMessage(e) + "' See hive log file for details.");
                    }
                    if (!HiveMetaStore.isRenameAllowed(olddb, db)) {
                        LOG.error("Alter Table operation for " + TableName.getQualified(catName, dbname, name) + "to new table = " + TableName.getQualified(catName, newDbName, newTblName) + " failed ");
                        throw new MetaException("Alter table not allowed for table " + TableName.getQualified(catName, dbname, name) + "to new table = " + TableName.getQualified(catName, newDbName, newTblName));
                    }
                }
            }
            if (isPartitionedTable) {
                String oldTblLocPath = srcPath.toUri().getPath();
                String newTblLocPath = dataWasMoved ? destPath.toUri().getPath() : null;
                // also the location field in partition
                parts = msdb.getPartitions(catName, dbname, name, -1);
                Multimap<Partition, ColumnStatistics> columnStatsNeedUpdated = ArrayListMultimap.create();
                for (Partition part : parts) {
                    String oldPartLoc = part.getSd().getLocation();
                    if (dataWasMoved && oldPartLoc.contains(oldTblLocPath)) {
                        URI oldUri = new Path(oldPartLoc).toUri();
                        String newPath = oldUri.getPath().replace(oldTblLocPath, newTblLocPath);
                        Path newPartLocPath = new Path(oldUri.getScheme(), oldUri.getAuthority(), newPath);
                        part.getSd().setLocation(newPartLocPath.toString());
                    }
                    part.setDbName(newDbName);
                    part.setTableName(newTblName);
                    List<ColumnStatistics> multiColStats = updateOrGetPartitionColumnStats(msdb, catName, dbname, name, part.getValues(), part.getSd().getCols(), oldt, part, null, null);
                    for (ColumnStatistics colStats : multiColStats) {
                        columnStatsNeedUpdated.put(part, colStats);
                    }
                }
                // Do not verify stats parameters on a partitioned table.
                msdb.alterTable(catName, dbname, name, newt, null);
                // alterPartition is only for changing the partition location in the table rename
                if (dataWasMoved) {
                    int partsToProcess = parts.size();
                    int partitionBatchSize = MetastoreConf.getIntVar(handler.getConf(), MetastoreConf.ConfVars.BATCH_RETRIEVE_MAX);
                    int batchStart = 0;
                    while (partsToProcess > 0) {
                        int batchEnd = Math.min(batchStart + partitionBatchSize, parts.size());
                        List<Partition> partBatch = parts.subList(batchStart, batchEnd);
                        int partBatchSize = partBatch.size();
                        partsToProcess -= partBatchSize;
                        batchStart += partBatchSize;
                        List<List<String>> partValues = new ArrayList<>(partBatchSize);
                        for (Partition part : partBatch) {
                            partValues.add(part.getValues());
                        }
                        msdb.alterPartitions(catName, newDbName, newTblName, partValues, partBatch, newt.getWriteId(), writeIdList);
                    }
                }
                Deadline.checkTimeout();
                for (Entry<Partition, ColumnStatistics> partColStats : columnStatsNeedUpdated.entries()) {
                    ColumnStatistics newPartColStats = partColStats.getValue();
                    newPartColStats.getStatsDesc().setDbName(newDbName);
                    newPartColStats.getStatsDesc().setTableName(newTblName);
                    msdb.updatePartitionColumnStatistics(newPartColStats, partColStats.getKey().getValues(), writeIdList, newt.getWriteId());
                }
            } else {
                alterTableUpdateTableColumnStats(msdb, oldt, newt, environmentContext, writeIdList, conf, null);
            }
        } else {
            // operations other than table rename
            if (MetaStoreServerUtils.requireCalStats(null, null, newt, environmentContext) && !isPartitionedTable) {
                Database db = msdb.getDatabase(catName, newDbName);
                assert (isReplicated == HMSHandler.isDbReplicationTarget(db));
                // Update table stats. For partitioned table, we update stats in alterPartition()
                MetaStoreServerUtils.updateTableStatsSlow(db, newt, wh, false, true, environmentContext);
            }
            if (isPartitionedTable) {
                // Currently only column related changes can be cascaded in alter table
                boolean runPartitionMetadataUpdate = (cascade && !MetaStoreServerUtils.areSameColumns(oldt.getSd().getCols(), newt.getSd().getCols()));
                // we may skip the update entirely if there are only new columns added
                runPartitionMetadataUpdate |= !cascade && !MetaStoreServerUtils.arePrefixColumns(oldt.getSd().getCols(), newt.getSd().getCols());
                boolean retainOnColRemoval = MetastoreConf.getBoolVar(handler.getConf(), MetastoreConf.ConfVars.COLSTATS_RETAIN_ON_COLUMN_REMOVAL);
                if (runPartitionMetadataUpdate) {
                    if (cascade || retainOnColRemoval) {
                        parts = msdb.getPartitions(catName, dbname, name, -1);
                        for (Partition part : parts) {
                            Partition oldPart = new Partition(part);
                            List<FieldSchema> oldCols = part.getSd().getCols();
                            part.getSd().setCols(newt.getSd().getCols());
                            List<ColumnStatistics> colStats = updateOrGetPartitionColumnStats(msdb, catName, dbname, name, part.getValues(), oldCols, oldt, part, null, null);
                            assert (colStats.isEmpty());
                            Deadline.checkTimeout();
                            if (cascade) {
                                msdb.alterPartition(catName, dbname, name, part.getValues(), part, writeIdList);
                            } else {
                                // update changed properties (stats)
                                oldPart.setParameters(part.getParameters());
                                msdb.alterPartition(catName, dbname, name, part.getValues(), oldPart, writeIdList);
                            }
                        }
                    } else {
                        // clear all column stats to prevent incorract behaviour in case same column is reintroduced
                        TableName tableName = new TableName(catName, dbname, name);
                        msdb.deleteAllPartitionColumnStatistics(tableName, writeIdList);
                    }
                    // Don't validate table-level stats for a partitoned table.
                    msdb.alterTable(catName, dbname, name, newt, null);
                } else {
                    LOG.warn("Alter table not cascaded to partitions.");
                    alterTableUpdateTableColumnStats(msdb, oldt, newt, environmentContext, writeIdList, conf, null);
                }
            } else {
                alterTableUpdateTableColumnStats(msdb, oldt, newt, environmentContext, writeIdList, conf, null);
            }
        }
        if (transactionalListeners != null && !transactionalListeners.isEmpty()) {
            txnAlterTableEventResponses = MetaStoreListenerNotifier.notifyEvent(transactionalListeners, EventMessage.EventType.ALTER_TABLE, new AlterTableEvent(oldt, newt, false, true, newt.getWriteId(), handler, isReplicated), environmentContext);
        }
        // commit the changes
        success = msdb.commitTransaction();
    } catch (InvalidObjectException e) {
        LOG.debug("Failed to get object from Metastore ", e);
        throw new InvalidOperationException("Unable to change partition or table." + " Check metastore logs for detailed stack." + e.getMessage());
    } catch (InvalidInputException e) {
        LOG.debug("Accessing Metastore failed due to invalid input ", e);
        throw new InvalidOperationException("Unable to change partition or table." + " Check metastore logs for detailed stack." + e.getMessage());
    } catch (NoSuchObjectException e) {
        LOG.debug("Object not found in metastore ", e);
        throw new InvalidOperationException("Unable to change partition or table. Object " + e.getMessage() + " does not exist." + " Check metastore logs for detailed stack.");
    } finally {
        if (success) {
            // If data location is changed in replication flow, then need to delete the old path.
            if (replDataLocationChanged) {
                assert (olddb != null);
                assert (oldt != null);
                Path deleteOldDataLoc = new Path(oldt.getSd().getLocation());
                boolean isSkipTrash = MetaStoreUtils.isSkipTrash(oldt.getParameters());
                try {
                    wh.deleteDir(deleteOldDataLoc, true, isSkipTrash, ReplChangeManager.shouldEnableCm(olddb, oldt));
                    LOG.info("Deleted the old data location: {} for the table: {}", deleteOldDataLoc, dbname + "." + name);
                } catch (MetaException ex) {
                    // Eat the exception as it doesn't affect the state of existing tables.
                    // Expect, user to manually drop this path when exception and so logging a warning.
                    LOG.warn("Unable to delete the old data location: {} for the table: {}", deleteOldDataLoc, dbname + "." + name);
                }
            }
        } else {
            LOG.error("Failed to alter table " + TableName.getQualified(catName, dbname, name));
            msdb.rollbackTransaction();
            if (!replDataLocationChanged && dataWasMoved) {
                try {
                    if (destFs.exists(destPath)) {
                        if (!destFs.rename(destPath, srcPath)) {
                            LOG.error("Failed to restore data from " + destPath + " to " + srcPath + " in alter table failure. Manual restore is needed.");
                        }
                    }
                } catch (IOException e) {
                    LOG.error("Failed to restore data from " + destPath + " to " + srcPath + " in alter table failure. Manual restore is needed.");
                }
            }
        }
    }
    if (!listeners.isEmpty()) {
        // I don't think event notifications in case of failures are necessary, but other HMS operations
        // make this call whether the event failed or succeeded. To make this behavior consistent,
        // this call is made for failed events also.
        MetaStoreListenerNotifier.notifyEvent(listeners, EventMessage.EventType.ALTER_TABLE, new AlterTableEvent(oldt, newt, false, success, newt.getWriteId(), handler, isReplicated), environmentContext, txnAlterTableEventResponses, msdb);
    }
}
Also used : FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema) ArrayList(java.util.ArrayList) URI(java.net.URI) AlterTableEvent(org.apache.hadoop.hive.metastore.events.AlterTableEvent) FileSystem(org.apache.hadoop.fs.FileSystem) InvalidOperationException(org.apache.hadoop.hive.metastore.api.InvalidOperationException) Database(org.apache.hadoop.hive.metastore.api.Database) List(java.util.List) ArrayList(java.util.ArrayList) LinkedList(java.util.LinkedList) InvalidObjectException(org.apache.hadoop.hive.metastore.api.InvalidObjectException) MetaException(org.apache.hadoop.hive.metastore.api.MetaException) Path(org.apache.hadoop.fs.Path) ColumnStatistics(org.apache.hadoop.hive.metastore.api.ColumnStatistics) Partition(org.apache.hadoop.hive.metastore.api.Partition) InvalidInputException(org.apache.hadoop.hive.metastore.api.InvalidInputException) Table(org.apache.hadoop.hive.metastore.api.Table) IOException(java.io.IOException) TableName(org.apache.hadoop.hive.common.TableName) NoSuchObjectException(org.apache.hadoop.hive.metastore.api.NoSuchObjectException)

Aggregations

TableName (org.apache.hadoop.hive.common.TableName)47 DDLWork (org.apache.hadoop.hive.ql.ddl.DDLWork)22 ArrayList (java.util.ArrayList)16 SemanticException (org.apache.hadoop.hive.ql.parse.SemanticException)15 Table (org.apache.hadoop.hive.ql.metadata.Table)14 FieldSchema (org.apache.hadoop.hive.metastore.api.FieldSchema)11 MetaException (org.apache.hadoop.hive.metastore.api.MetaException)11 HiveTableName (org.apache.hadoop.hive.ql.parse.HiveTableName)10 HashMap (java.util.HashMap)9 HiveException (org.apache.hadoop.hive.ql.metadata.HiveException)9 HashSet (java.util.HashSet)8 IOException (java.io.IOException)7 List (java.util.List)7 Path (org.apache.hadoop.fs.Path)7 SQLCheckConstraint (org.apache.hadoop.hive.metastore.api.SQLCheckConstraint)7 SQLDefaultConstraint (org.apache.hadoop.hive.metastore.api.SQLDefaultConstraint)7 SQLNotNullConstraint (org.apache.hadoop.hive.metastore.api.SQLNotNullConstraint)7 SQLUniqueConstraint (org.apache.hadoop.hive.metastore.api.SQLUniqueConstraint)7 ASTNode (org.apache.hadoop.hive.ql.parse.ASTNode)7 SourceTable (org.apache.hadoop.hive.metastore.api.SourceTable)6