Search in sources :

Example 1 with AccumuloException

use of org.apache.accumulo.core.client.AccumuloException in project hive by apache.

the class HiveAccumuloTableInputFormat method getSplits.

@Override
public InputSplit[] getSplits(JobConf jobConf, int numSplits) throws IOException {
    final AccumuloConnectionParameters accumuloParams = new AccumuloConnectionParameters(jobConf);
    final Instance instance = accumuloParams.getInstance();
    final ColumnMapper columnMapper;
    try {
        columnMapper = getColumnMapper(jobConf);
    } catch (TooManyAccumuloColumnsException e) {
        throw new IOException(e);
    }
    JobContext context = ShimLoader.getHadoopShims().newJobContext(Job.getInstance(jobConf));
    Path[] tablePaths = FileInputFormat.getInputPaths(context);
    try {
        UserGroupInformation ugi = UserGroupInformation.getCurrentUser();
        final Connector connector;
        // Need to get a Connector so we look up the user's authorizations if not otherwise specified
        if (accumuloParams.useSasl() && !ugi.hasKerberosCredentials()) {
            // In a YARN/Tez job, don't have the Kerberos credentials anymore, use the delegation token
            AuthenticationToken token = ConfiguratorBase.getAuthenticationToken(AccumuloInputFormat.class, jobConf);
            // Convert the stub from the configuration back into a normal Token
            // More reflection to support 1.6
            token = helper.unwrapAuthenticationToken(jobConf, token);
            connector = instance.getConnector(accumuloParams.getAccumuloUserName(), token);
        } else {
            // Still in the local JVM, use the username+password or Kerberos credentials
            connector = accumuloParams.getConnector(instance);
        }
        final List<ColumnMapping> columnMappings = columnMapper.getColumnMappings();
        final List<IteratorSetting> iterators = predicateHandler.getIterators(jobConf, columnMapper);
        final Collection<Range> ranges = predicateHandler.getRanges(jobConf, columnMapper);
        // We don't want that.
        if (null != ranges && ranges.isEmpty()) {
            return new InputSplit[0];
        }
        // Set the relevant information in the Configuration for the AccumuloInputFormat
        configure(jobConf, instance, connector, accumuloParams, columnMapper, iterators, ranges);
        int numColumns = columnMappings.size();
        List<Integer> readColIds = ColumnProjectionUtils.getReadColumnIDs(jobConf);
        // Sanity check
        if (numColumns < readColIds.size())
            throw new IOException("Number of column mappings (" + numColumns + ")" + " numbers less than the hive table columns. (" + readColIds.size() + ")");
        // get splits from Accumulo
        InputSplit[] splits = accumuloInputFormat.getSplits(jobConf, numSplits);
        HiveAccumuloSplit[] hiveSplits = new HiveAccumuloSplit[splits.length];
        for (int i = 0; i < splits.length; i++) {
            RangeInputSplit ris = (RangeInputSplit) splits[i];
            hiveSplits[i] = new HiveAccumuloSplit(ris, tablePaths[0]);
        }
        return hiveSplits;
    } catch (AccumuloException e) {
        log.error("Could not configure AccumuloInputFormat", e);
        throw new IOException(StringUtils.stringifyException(e));
    } catch (AccumuloSecurityException e) {
        log.error("Could not configure AccumuloInputFormat", e);
        throw new IOException(StringUtils.stringifyException(e));
    } catch (SerDeException e) {
        log.error("Could not configure AccumuloInputFormat", e);
        throw new IOException(StringUtils.stringifyException(e));
    }
}
Also used : Connector(org.apache.accumulo.core.client.Connector) AuthenticationToken(org.apache.accumulo.core.client.security.tokens.AuthenticationToken) MockInstance(org.apache.accumulo.core.client.mock.MockInstance) Instance(org.apache.accumulo.core.client.Instance) RangeInputSplit(org.apache.accumulo.core.client.mapred.RangeInputSplit) AccumuloSecurityException(org.apache.accumulo.core.client.AccumuloSecurityException) JobContext(org.apache.hadoop.mapreduce.JobContext) RangeInputSplit(org.apache.accumulo.core.client.mapred.RangeInputSplit) InputSplit(org.apache.hadoop.mapred.InputSplit) HiveAccumuloMapColumnMapping(org.apache.hadoop.hive.accumulo.columns.HiveAccumuloMapColumnMapping) ColumnMapping(org.apache.hadoop.hive.accumulo.columns.ColumnMapping) HiveAccumuloColumnMapping(org.apache.hadoop.hive.accumulo.columns.HiveAccumuloColumnMapping) SerDeException(org.apache.hadoop.hive.serde2.SerDeException) UserGroupInformation(org.apache.hadoop.security.UserGroupInformation) Path(org.apache.hadoop.fs.Path) AccumuloException(org.apache.accumulo.core.client.AccumuloException) IOException(java.io.IOException) TooManyAccumuloColumnsException(org.apache.hadoop.hive.accumulo.serde.TooManyAccumuloColumnsException) Range(org.apache.accumulo.core.data.Range) IteratorSetting(org.apache.accumulo.core.client.IteratorSetting) AccumuloConnectionParameters(org.apache.hadoop.hive.accumulo.AccumuloConnectionParameters) ColumnMapper(org.apache.hadoop.hive.accumulo.columns.ColumnMapper)

Example 2 with AccumuloException

use of org.apache.accumulo.core.client.AccumuloException in project hive by apache.

the class HiveAccumuloTableOutputFormat method configureAccumuloOutputFormat.

protected void configureAccumuloOutputFormat(JobConf job) throws IOException {
    AccumuloConnectionParameters cnxnParams = getConnectionParams(job);
    final String tableName = job.get(AccumuloSerDeParameters.TABLE_NAME);
    // Make sure we actually go the table name
    Preconditions.checkNotNull(tableName, "Expected Accumulo table name to be provided in job configuration");
    // Set the necessary Accumulo information
    try {
        if (cnxnParams.useMockInstance()) {
            setMockInstanceWithErrorChecking(job, cnxnParams.getAccumuloInstanceName());
        } else {
            // Accumulo instance name with ZK quorum
            setZooKeeperInstanceWithErrorChecking(job, cnxnParams.getAccumuloInstanceName(), cnxnParams.getZooKeepers(), cnxnParams.useSasl());
        }
        // The AccumuloOutputFormat will look for it there.
        if (cnxnParams.useSasl()) {
            UserGroupInformation ugi = getCurrentUser();
            if (!hasKerberosCredentials(ugi)) {
                getHelper().addTokenFromUserToJobConf(ugi, job);
            } else {
                // Still in the local JVM, can use Kerberos credentials
                try {
                    Connector connector = cnxnParams.getConnector();
                    AuthenticationToken token = getHelper().getDelegationToken(connector);
                    // Send the DelegationToken down to the Configuration for Accumulo to use
                    setConnectorInfoWithErrorChecking(job, cnxnParams.getAccumuloUserName(), token);
                    // Convert the Accumulo token in a Hadoop token
                    Token<? extends TokenIdentifier> accumuloToken = getHelper().getHadoopToken(token);
                    log.info("Adding Hadoop Token for Accumulo to Job's Credentials");
                    // Add the Hadoop token to the JobConf
                    getHelper().mergeTokenIntoJobConf(job, accumuloToken);
                    // Make sure the UGI contains the token too for good measure
                    if (!ugi.addToken(accumuloToken)) {
                        throw new IOException("Failed to add Accumulo Token to UGI");
                    }
                } catch (AccumuloException | AccumuloSecurityException e) {
                    throw new IOException("Failed to acquire Accumulo DelegationToken", e);
                }
            }
        } else {
            setConnectorInfoWithErrorChecking(job, cnxnParams.getAccumuloUserName(), new PasswordToken(cnxnParams.getAccumuloPassword()));
        }
        // Set the table where we're writing this data
        setDefaultAccumuloTableName(job, tableName);
    } catch (AccumuloSecurityException e) {
        log.error("Could not connect to Accumulo with provided credentials", e);
        throw new IOException(e);
    }
}
Also used : Connector(org.apache.accumulo.core.client.Connector) AccumuloException(org.apache.accumulo.core.client.AccumuloException) PasswordToken(org.apache.accumulo.core.client.security.tokens.PasswordToken) AuthenticationToken(org.apache.accumulo.core.client.security.tokens.AuthenticationToken) AccumuloConnectionParameters(org.apache.hadoop.hive.accumulo.AccumuloConnectionParameters) AccumuloSecurityException(org.apache.accumulo.core.client.AccumuloSecurityException) IOException(java.io.IOException) UserGroupInformation(org.apache.hadoop.security.UserGroupInformation)

Example 3 with AccumuloException

use of org.apache.accumulo.core.client.AccumuloException in project YCSB by brianfrankcooper.

the class AccumuloClient method init.

@Override
public void init() throws DBException {
    colFam = new Text(getProperties().getProperty("accumulo.columnFamily"));
    inst = new ZooKeeperInstance(getProperties().getProperty("accumulo.instanceName"), getProperties().getProperty("accumulo.zooKeepers"));
    try {
        String principal = getProperties().getProperty("accumulo.username");
        AuthenticationToken token = new PasswordToken(getProperties().getProperty("accumulo.password"));
        connector = inst.getConnector(principal, token);
    } catch (AccumuloException e) {
        throw new DBException(e);
    } catch (AccumuloSecurityException e) {
        throw new DBException(e);
    }
    if (!(getProperties().getProperty("accumulo.pcFlag", "none").equals("none"))) {
        System.err.println("Sorry, the ZK based producer/consumer implementation has been removed. " + "Please see YCSB issue #416 for work on adding a general solution to coordinated work.");
    }
}
Also used : AccumuloException(org.apache.accumulo.core.client.AccumuloException) DBException(com.yahoo.ycsb.DBException) PasswordToken(org.apache.accumulo.core.client.security.tokens.PasswordToken) AuthenticationToken(org.apache.accumulo.core.client.security.tokens.AuthenticationToken) Text(org.apache.hadoop.io.Text) AccumuloSecurityException(org.apache.accumulo.core.client.AccumuloSecurityException) ZooKeeperInstance(org.apache.accumulo.core.client.ZooKeeperInstance)

Example 4 with AccumuloException

use of org.apache.accumulo.core.client.AccumuloException in project Gaffer by gchq.

the class CoreKeyGroupByAggregatorIteratorTest method shouldPartiallyAggregateColumnQualifierOverCQ1GroupBy.

public void shouldPartiallyAggregateColumnQualifierOverCQ1GroupBy(final AccumuloStore store, final AccumuloElementConverter elementConverter) throws StoreException, AccumuloElementConversionException {
    final String visibilityString = "public";
    try {
        // Create edge
        final Edge edge = new Edge(TestGroups.EDGE);
        edge.setSource("1");
        edge.setDestination("2");
        edge.setDirected(true);
        edge.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        edge.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 1);
        edge.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge.putProperty(AccumuloPropertyNames.COUNT, 1);
        final Edge edge2 = new Edge(TestGroups.EDGE);
        edge2.setSource("1");
        edge2.setDestination("2");
        edge2.setDirected(true);
        edge2.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        edge2.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 2);
        edge2.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge2.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge2.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge2.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge2.putProperty(AccumuloPropertyNames.COUNT, 1);
        final Edge edge3 = new Edge(TestGroups.EDGE);
        edge3.setSource("1");
        edge3.setDestination("2");
        edge3.setDirected(true);
        edge3.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        edge3.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 3);
        edge3.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge3.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge3.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge3.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge3.putProperty(AccumuloPropertyNames.COUNT, 1);
        //THIS EDGE WILL BE REDUCED MEANING ITS CQ (columnQualifier) will only occur once because its key is equal.
        final Edge edge4 = new Edge(TestGroups.EDGE);
        edge4.setSource("1");
        edge4.setDestination("2");
        edge4.setDirected(true);
        edge4.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 2);
        edge4.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 4);
        edge4.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge4.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge4.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge4.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge4.putProperty(AccumuloPropertyNames.COUNT, 2);
        final Edge edge5 = new Edge(TestGroups.EDGE);
        edge5.setSource("1");
        edge5.setDestination("2");
        edge5.setDirected(true);
        edge5.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 3);
        edge5.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 5);
        edge5.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge5.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge5.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge5.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge5.putProperty(AccumuloPropertyNames.COUNT, 10);
        final Edge edge6 = new Edge(TestGroups.EDGE);
        edge6.setSource("1");
        edge6.setDestination("2");
        edge6.setDirected(true);
        edge6.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 4);
        edge6.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 6);
        edge6.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge6.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge6.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge6.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge6.putProperty(AccumuloPropertyNames.COUNT, 5);
        // Accumulo key
        final Key key = elementConverter.getKeysFromEdge(edge).getFirst();
        final Key key2 = elementConverter.getKeysFromEdge(edge2).getFirst();
        final Key key3 = elementConverter.getKeysFromEdge(edge3).getFirst();
        final Key key4 = elementConverter.getKeysFromEdge(edge4).getFirst();
        final Key key5 = elementConverter.getKeysFromEdge(edge5).getFirst();
        final Key key6 = elementConverter.getKeysFromEdge(edge6).getFirst();
        // Accumulo values
        final Value value1 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge.getProperties());
        final Value value2 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge2.getProperties());
        final Value value3 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge3.getProperties());
        final Value value4 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge4.getProperties());
        final Value value5 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge5.getProperties());
        final Value value6 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge6.getProperties());
        // Create mutation
        final Mutation m1 = new Mutation(key.getRow());
        m1.put(key.getColumnFamily(), key.getColumnQualifier(), new ColumnVisibility(key.getColumnVisibility()), key.getTimestamp(), value1);
        final Mutation m2 = new Mutation(key2.getRow());
        m2.put(key2.getColumnFamily(), key2.getColumnQualifier(), new ColumnVisibility(key2.getColumnVisibility()), key2.getTimestamp(), value2);
        final Mutation m3 = new Mutation(key.getRow());
        m3.put(key3.getColumnFamily(), key3.getColumnQualifier(), new ColumnVisibility(key3.getColumnVisibility()), key3.getTimestamp(), value3);
        final Mutation m4 = new Mutation(key.getRow());
        m4.put(key4.getColumnFamily(), key4.getColumnQualifier(), new ColumnVisibility(key4.getColumnVisibility()), key4.getTimestamp(), value4);
        final Mutation m5 = new Mutation(key.getRow());
        m5.put(key5.getColumnFamily(), key5.getColumnQualifier(), new ColumnVisibility(key5.getColumnVisibility()), key5.getTimestamp(), value5);
        final Mutation m6 = new Mutation(key.getRow());
        m6.put(key6.getColumnFamily(), key6.getColumnQualifier(), new ColumnVisibility(key6.getColumnVisibility()), key6.getTimestamp(), value6);
        // Write mutation
        final BatchWriterConfig writerConfig = new BatchWriterConfig();
        writerConfig.setMaxMemory(1000000L);
        writerConfig.setMaxLatency(1000L, TimeUnit.MILLISECONDS);
        writerConfig.setMaxWriteThreads(1);
        final BatchWriter writer = store.getConnection().createBatchWriter(store.getProperties().getTable(), writerConfig);
        writer.addMutation(m1);
        writer.addMutation(m2);
        writer.addMutation(m3);
        writer.addMutation(m4);
        writer.addMutation(m5);
        writer.addMutation(m6);
        writer.close();
        Edge expectedEdge1 = new Edge(TestGroups.EDGE);
        expectedEdge1.setSource("1");
        expectedEdge1.setDestination("2");
        expectedEdge1.setDirected(true);
        expectedEdge1.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        expectedEdge1.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 6);
        expectedEdge1.putProperty(AccumuloPropertyNames.COUNT, 3);
        expectedEdge1.putProperty(AccumuloPropertyNames.PROP_1, 0);
        expectedEdge1.putProperty(AccumuloPropertyNames.PROP_2, 0);
        expectedEdge1.putProperty(AccumuloPropertyNames.PROP_3, 0);
        expectedEdge1.putProperty(AccumuloPropertyNames.PROP_4, 0);
        Edge expectedEdge2 = new Edge(TestGroups.EDGE);
        expectedEdge2.setSource("1");
        expectedEdge2.setDestination("2");
        expectedEdge2.setDirected(true);
        expectedEdge2.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 2);
        expectedEdge2.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 4);
        expectedEdge2.putProperty(AccumuloPropertyNames.COUNT, 2);
        expectedEdge2.putProperty(AccumuloPropertyNames.PROP_1, 0);
        expectedEdge2.putProperty(AccumuloPropertyNames.PROP_2, 0);
        expectedEdge2.putProperty(AccumuloPropertyNames.PROP_3, 0);
        expectedEdge2.putProperty(AccumuloPropertyNames.PROP_4, 0);
        Edge expectedEdge3 = new Edge(TestGroups.EDGE);
        expectedEdge3.setSource("1");
        expectedEdge3.setDestination("2");
        expectedEdge3.setDirected(true);
        expectedEdge3.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 3);
        expectedEdge3.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 5);
        expectedEdge3.putProperty(AccumuloPropertyNames.COUNT, 10);
        expectedEdge3.putProperty(AccumuloPropertyNames.PROP_1, 0);
        expectedEdge3.putProperty(AccumuloPropertyNames.PROP_2, 0);
        expectedEdge3.putProperty(AccumuloPropertyNames.PROP_3, 0);
        expectedEdge3.putProperty(AccumuloPropertyNames.PROP_4, 0);
        Edge expectedEdge4 = new Edge(TestGroups.EDGE);
        expectedEdge4.setSource("1");
        expectedEdge4.setDestination("2");
        expectedEdge4.setDirected(true);
        expectedEdge4.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 4);
        expectedEdge4.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 6);
        expectedEdge4.putProperty(AccumuloPropertyNames.COUNT, 5);
        expectedEdge4.putProperty(AccumuloPropertyNames.PROP_1, 0);
        expectedEdge4.putProperty(AccumuloPropertyNames.PROP_2, 0);
        expectedEdge4.putProperty(AccumuloPropertyNames.PROP_3, 0);
        expectedEdge4.putProperty(AccumuloPropertyNames.PROP_4, 0);
        // Read data back and check we get one merged element
        final Authorizations authorizations = new Authorizations(visibilityString);
        final Scanner scanner = store.getConnection().createScanner(store.getProperties().getTable(), authorizations);
        final IteratorSetting iteratorSetting = new IteratorSettingBuilder(AccumuloStoreConstants.COLUMN_QUALIFIER_AGGREGATOR_ITERATOR_PRIORITY, "KeyCombiner", CoreKeyGroupByAggregatorIterator.class).all().view(new View.Builder().edge(TestGroups.EDGE, new ViewElementDefinition.Builder().groupBy(AccumuloPropertyNames.COLUMN_QUALIFIER).build()).build()).schema(store.getSchema()).keyConverter(store.getKeyPackage().getKeyConverter()).build();
        scanner.addScanIterator(iteratorSetting);
        final Iterator<Entry<Key, Value>> it = scanner.iterator();
        Entry<Key, Value> entry = it.next();
        Element readEdge = elementConverter.getFullElement(entry.getKey(), entry.getValue());
        assertEquals(expectedEdge1, readEdge);
        assertEquals(1, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER));
        assertEquals(6, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2));
        assertEquals(3, readEdge.getProperty(AccumuloPropertyNames.COUNT));
        entry = it.next();
        readEdge = elementConverter.getFullElement(entry.getKey(), entry.getValue());
        assertEquals(expectedEdge2, readEdge);
        assertEquals(2, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER));
        assertEquals(4, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2));
        assertEquals(2, readEdge.getProperty(AccumuloPropertyNames.COUNT));
        entry = it.next();
        readEdge = elementConverter.getFullElement(entry.getKey(), entry.getValue());
        assertEquals(expectedEdge3, readEdge);
        assertEquals(3, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER));
        assertEquals(5, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2));
        assertEquals(10, readEdge.getProperty(AccumuloPropertyNames.COUNT));
        entry = it.next();
        readEdge = elementConverter.getFullElement(entry.getKey(), entry.getValue());
        assertEquals(expectedEdge4, readEdge);
        assertEquals(4, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER));
        assertEquals(6, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2));
        assertEquals(5, readEdge.getProperty(AccumuloPropertyNames.COUNT));
        if (it.hasNext()) {
            fail("Additional row found.");
        }
    } catch (AccumuloException | TableNotFoundException e) {
        fail(this.getClass().getSimpleName() + " failed with exception: " + e);
    }
}
Also used : Scanner(org.apache.accumulo.core.client.Scanner) AccumuloException(org.apache.accumulo.core.client.AccumuloException) Authorizations(org.apache.accumulo.core.security.Authorizations) IteratorSettingBuilder(uk.gov.gchq.gaffer.accumulostore.utils.IteratorSettingBuilder) Element(uk.gov.gchq.gaffer.data.element.Element) IteratorSettingBuilder(uk.gov.gchq.gaffer.accumulostore.utils.IteratorSettingBuilder) TableNotFoundException(org.apache.accumulo.core.client.TableNotFoundException) Entry(java.util.Map.Entry) IteratorSetting(org.apache.accumulo.core.client.IteratorSetting) Value(org.apache.accumulo.core.data.Value) BatchWriterConfig(org.apache.accumulo.core.client.BatchWriterConfig) Mutation(org.apache.accumulo.core.data.Mutation) ColumnVisibility(org.apache.accumulo.core.security.ColumnVisibility) BatchWriter(org.apache.accumulo.core.client.BatchWriter) Edge(uk.gov.gchq.gaffer.data.element.Edge) Key(org.apache.accumulo.core.data.Key)

Example 5 with AccumuloException

use of org.apache.accumulo.core.client.AccumuloException in project Gaffer by gchq.

the class CoreKeyGroupByAggregatorIteratorTest method shouldAggregatePropertiesOnlyWhenGroupByIsSetToCQ1CQ2.

public void shouldAggregatePropertiesOnlyWhenGroupByIsSetToCQ1CQ2(final AccumuloStore store, final AccumuloElementConverter elementConverter) throws StoreException, AccumuloElementConversionException {
    final String visibilityString = "public";
    try {
        // Create edge
        final Edge edge = new Edge(TestGroups.EDGE);
        edge.setSource("1");
        edge.setDestination("2");
        edge.setDirected(true);
        edge.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        edge.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 1);
        edge.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge.putProperty(AccumuloPropertyNames.COUNT, 1);
        final Edge edge2 = new Edge(TestGroups.EDGE);
        edge2.setSource("1");
        edge2.setDestination("2");
        edge2.setDirected(true);
        edge2.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        edge2.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 1);
        edge2.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge2.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge2.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge2.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge2.putProperty(AccumuloPropertyNames.COUNT, 1);
        final Edge edge3 = new Edge(TestGroups.EDGE);
        edge3.setSource("1");
        edge3.setDestination("2");
        edge3.setDirected(true);
        edge3.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        edge3.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 1);
        edge3.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge3.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge3.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge3.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge3.putProperty(AccumuloPropertyNames.COUNT, 1);
        //THIS EDGE WILL BE REDUCED MEANING ITS CQ (columnQualifier) will only occur once because its key is equal.
        final Edge edge4 = new Edge(TestGroups.EDGE);
        edge4.setSource("1");
        edge4.setDestination("2");
        edge4.setDirected(true);
        edge4.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        edge4.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 4);
        edge4.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge4.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge4.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge4.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge4.putProperty(AccumuloPropertyNames.COUNT, 2);
        final Edge edge5 = new Edge(TestGroups.EDGE);
        edge5.setSource("1");
        edge5.setDestination("2");
        edge5.setDirected(true);
        edge5.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 3);
        edge5.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 5);
        edge5.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge5.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge5.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge5.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge5.putProperty(AccumuloPropertyNames.COUNT, 10);
        final Edge edge6 = new Edge(TestGroups.EDGE);
        edge6.setSource("1");
        edge6.setDestination("2");
        edge6.setDirected(true);
        edge6.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 3);
        edge6.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 5);
        edge6.putProperty(AccumuloPropertyNames.PROP_1, 0);
        edge6.putProperty(AccumuloPropertyNames.PROP_2, 0);
        edge6.putProperty(AccumuloPropertyNames.PROP_3, 0);
        edge6.putProperty(AccumuloPropertyNames.PROP_4, 0);
        edge6.putProperty(AccumuloPropertyNames.COUNT, 5);
        // Accumulo key
        final Key key = elementConverter.getKeysFromEdge(edge).getFirst();
        final Key key2 = elementConverter.getKeysFromEdge(edge2).getFirst();
        final Key key3 = elementConverter.getKeysFromEdge(edge3).getFirst();
        final Key key4 = elementConverter.getKeysFromEdge(edge4).getFirst();
        final Key key5 = elementConverter.getKeysFromEdge(edge5).getFirst();
        final Key key6 = elementConverter.getKeysFromEdge(edge6).getFirst();
        // Accumulo values
        final Value value1 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge.getProperties());
        final Value value2 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge2.getProperties());
        final Value value3 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge3.getProperties());
        final Value value4 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge4.getProperties());
        final Value value5 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge5.getProperties());
        final Value value6 = elementConverter.getValueFromProperties(TestGroups.EDGE, edge6.getProperties());
        // Create mutation
        final Mutation m1 = new Mutation(key.getRow());
        m1.put(key.getColumnFamily(), key.getColumnQualifier(), new ColumnVisibility(key.getColumnVisibility()), key.getTimestamp(), value1);
        final Mutation m2 = new Mutation(key2.getRow());
        m2.put(key2.getColumnFamily(), key2.getColumnQualifier(), new ColumnVisibility(key2.getColumnVisibility()), key2.getTimestamp(), value2);
        final Mutation m3 = new Mutation(key.getRow());
        m3.put(key3.getColumnFamily(), key3.getColumnQualifier(), new ColumnVisibility(key3.getColumnVisibility()), key3.getTimestamp(), value3);
        final Mutation m4 = new Mutation(key.getRow());
        m4.put(key4.getColumnFamily(), key4.getColumnQualifier(), new ColumnVisibility(key4.getColumnVisibility()), key4.getTimestamp(), value4);
        final Mutation m5 = new Mutation(key.getRow());
        m5.put(key5.getColumnFamily(), key5.getColumnQualifier(), new ColumnVisibility(key5.getColumnVisibility()), key5.getTimestamp(), value5);
        final Mutation m6 = new Mutation(key.getRow());
        m6.put(key6.getColumnFamily(), key6.getColumnQualifier(), new ColumnVisibility(key6.getColumnVisibility()), key6.getTimestamp(), value6);
        // Write mutation
        final BatchWriterConfig writerConfig = new BatchWriterConfig();
        writerConfig.setMaxMemory(1000000L);
        writerConfig.setMaxLatency(1000L, TimeUnit.MILLISECONDS);
        writerConfig.setMaxWriteThreads(1);
        final BatchWriter writer = store.getConnection().createBatchWriter(store.getProperties().getTable(), writerConfig);
        writer.addMutation(m1);
        writer.addMutation(m2);
        writer.addMutation(m3);
        writer.addMutation(m4);
        writer.addMutation(m5);
        writer.addMutation(m6);
        writer.close();
        Edge expectedEdge1 = new Edge(TestGroups.EDGE);
        expectedEdge1.setSource("1");
        expectedEdge1.setDestination("2");
        expectedEdge1.setDirected(true);
        expectedEdge1.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        expectedEdge1.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 1);
        expectedEdge1.putProperty(AccumuloPropertyNames.COUNT, 3);
        expectedEdge1.putProperty(AccumuloPropertyNames.PROP_1, 0);
        expectedEdge1.putProperty(AccumuloPropertyNames.PROP_2, 0);
        expectedEdge1.putProperty(AccumuloPropertyNames.PROP_3, 0);
        expectedEdge1.putProperty(AccumuloPropertyNames.PROP_4, 0);
        Edge expectedEdge2 = new Edge(TestGroups.EDGE);
        expectedEdge2.setSource("1");
        expectedEdge2.setDestination("2");
        expectedEdge2.setDirected(true);
        expectedEdge2.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 1);
        expectedEdge2.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 4);
        expectedEdge2.putProperty(AccumuloPropertyNames.COUNT, 2);
        expectedEdge2.putProperty(AccumuloPropertyNames.PROP_1, 0);
        expectedEdge2.putProperty(AccumuloPropertyNames.PROP_2, 0);
        expectedEdge2.putProperty(AccumuloPropertyNames.PROP_3, 0);
        expectedEdge2.putProperty(AccumuloPropertyNames.PROP_4, 0);
        Edge expectedEdge3 = new Edge(TestGroups.EDGE);
        expectedEdge3.setSource("1");
        expectedEdge3.setDestination("2");
        expectedEdge3.setDirected(true);
        expectedEdge3.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER, 3);
        expectedEdge3.putProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2, 5);
        expectedEdge3.putProperty(AccumuloPropertyNames.COUNT, 15);
        expectedEdge3.putProperty(AccumuloPropertyNames.PROP_1, 0);
        expectedEdge3.putProperty(AccumuloPropertyNames.PROP_2, 0);
        expectedEdge3.putProperty(AccumuloPropertyNames.PROP_3, 0);
        expectedEdge3.putProperty(AccumuloPropertyNames.PROP_4, 0);
        // Read data back and check we get one merged element
        final Authorizations authorizations = new Authorizations(visibilityString);
        final Scanner scanner = store.getConnection().createScanner(store.getProperties().getTable(), authorizations);
        final IteratorSetting iteratorSetting = new IteratorSettingBuilder(AccumuloStoreConstants.COLUMN_QUALIFIER_AGGREGATOR_ITERATOR_PRIORITY, "KeyCombiner", CoreKeyGroupByAggregatorIterator.class).all().view(new View.Builder().edge(TestGroups.EDGE, new ViewElementDefinition.Builder().groupBy(AccumuloPropertyNames.COLUMN_QUALIFIER, AccumuloPropertyNames.COLUMN_QUALIFIER_2).build()).build()).schema(store.getSchema()).keyConverter(store.getKeyPackage().getKeyConverter()).build();
        scanner.addScanIterator(iteratorSetting);
        final Iterator<Entry<Key, Value>> it = scanner.iterator();
        Entry<Key, Value> entry = it.next();
        Element readEdge = elementConverter.getFullElement(entry.getKey(), entry.getValue());
        assertEquals(expectedEdge1, readEdge);
        assertEquals(1, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER));
        assertEquals(1, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2));
        assertEquals(3, readEdge.getProperty(AccumuloPropertyNames.COUNT));
        entry = it.next();
        readEdge = elementConverter.getFullElement(entry.getKey(), entry.getValue());
        assertEquals(expectedEdge2, readEdge);
        assertEquals(1, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER));
        assertEquals(4, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2));
        assertEquals(2, readEdge.getProperty(AccumuloPropertyNames.COUNT));
        entry = it.next();
        readEdge = elementConverter.getFullElement(entry.getKey(), entry.getValue());
        assertEquals(expectedEdge3, readEdge);
        assertEquals(3, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER));
        assertEquals(5, readEdge.getProperty(AccumuloPropertyNames.COLUMN_QUALIFIER_2));
        assertEquals(15, readEdge.getProperty(AccumuloPropertyNames.COUNT));
        if (it.hasNext()) {
            fail("Additional row found.");
        }
    } catch (AccumuloException | TableNotFoundException e) {
        fail(this.getClass().getSimpleName() + " failed with exception: " + e);
    }
}
Also used : Scanner(org.apache.accumulo.core.client.Scanner) AccumuloException(org.apache.accumulo.core.client.AccumuloException) Authorizations(org.apache.accumulo.core.security.Authorizations) IteratorSettingBuilder(uk.gov.gchq.gaffer.accumulostore.utils.IteratorSettingBuilder) Element(uk.gov.gchq.gaffer.data.element.Element) IteratorSettingBuilder(uk.gov.gchq.gaffer.accumulostore.utils.IteratorSettingBuilder) TableNotFoundException(org.apache.accumulo.core.client.TableNotFoundException) Entry(java.util.Map.Entry) IteratorSetting(org.apache.accumulo.core.client.IteratorSetting) Value(org.apache.accumulo.core.data.Value) BatchWriterConfig(org.apache.accumulo.core.client.BatchWriterConfig) Mutation(org.apache.accumulo.core.data.Mutation) ColumnVisibility(org.apache.accumulo.core.security.ColumnVisibility) BatchWriter(org.apache.accumulo.core.client.BatchWriter) Edge(uk.gov.gchq.gaffer.data.element.Edge) Key(org.apache.accumulo.core.data.Key)

Aggregations

AccumuloException (org.apache.accumulo.core.client.AccumuloException)21 TableNotFoundException (org.apache.accumulo.core.client.TableNotFoundException)15 AccumuloSecurityException (org.apache.accumulo.core.client.AccumuloSecurityException)14 IOException (java.io.IOException)7 Entry (java.util.Map.Entry)7 BatchWriter (org.apache.accumulo.core.client.BatchWriter)7 BatchWriterConfig (org.apache.accumulo.core.client.BatchWriterConfig)7 IteratorSetting (org.apache.accumulo.core.client.IteratorSetting)7 Key (org.apache.accumulo.core.data.Key)7 Mutation (org.apache.accumulo.core.data.Mutation)7 Value (org.apache.accumulo.core.data.Value)7 Authorizations (org.apache.accumulo.core.security.Authorizations)7 ColumnVisibility (org.apache.accumulo.core.security.ColumnVisibility)7 Edge (uk.gov.gchq.gaffer.data.element.Edge)7 Element (uk.gov.gchq.gaffer.data.element.Element)7 Scanner (org.apache.accumulo.core.client.Scanner)6 Text (org.apache.hadoop.io.Text)6 IteratorSettingBuilder (uk.gov.gchq.gaffer.accumulostore.utils.IteratorSettingBuilder)6 Connector (org.apache.accumulo.core.client.Connector)5 PasswordToken (org.apache.accumulo.core.client.security.tokens.PasswordToken)4