Examples with SparkRDDWriteClient org.apache.hudi.client.SparkRDDWriteClient used on opensource projects

Example 41 with SparkRDDWriteClient

use of org.apache.hudi.client.SparkRDDWriteClient in project hudi by apache.

the class TestSparkHoodieHBaseIndex method testSimpleTagLocationAndUpdateWithRollback.

@Test
public void testSimpleTagLocationAndUpdateWithRollback() throws Exception {
    // Load to memory
    HoodieWriteConfig config = getConfigBuilder(100, false, false).withRollbackUsingMarkers(false).build();
    SparkHoodieHBaseIndex index = new SparkHoodieHBaseIndex(config);
    SparkRDDWriteClient writeClient = getHoodieWriteClient(config);
    final String newCommitTime = writeClient.startCommit();
    final int numRecords = 10;
    List<HoodieRecord> records = dataGen.generateInserts(newCommitTime, numRecords);
    JavaRDD<HoodieRecord> writeRecords = jsc().parallelize(records, 1);
    metaClient = HoodieTableMetaClient.reload(metaClient);
    // Insert 200 records
    JavaRDD<WriteStatus> writeStatues = writeClient.upsert(writeRecords, newCommitTime);
    assertNoWriteErrors(writeStatues.collect());
    // commit this upsert
    writeClient.commit(newCommitTime, writeStatues);
    HoodieTable hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    // Now tagLocation for these records, hbaseIndex should tag them
    List<HoodieRecord> records2 = tagLocation(index, writeRecords, hoodieTable).collect();
    assertEquals(numRecords, records2.stream().filter(HoodieRecord::isCurrentLocationKnown).count());
    // check tagged records are tagged with correct fileIds
    List<String> fileIds = writeStatues.map(WriteStatus::getFileId).collect();
    assertEquals(0, records2.stream().filter(record -> record.getCurrentLocation().getFileId() == null).count());
    List<String> taggedFileIds = records2.stream().map(record -> record.getCurrentLocation().getFileId()).distinct().collect(Collectors.toList());
    // both lists should match
    assertTrue(taggedFileIds.containsAll(fileIds) && fileIds.containsAll(taggedFileIds));
    // Rollback the last commit
    writeClient.rollback(newCommitTime);
    hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    // Now tagLocation for these records, hbaseIndex should not tag them since it was a rolled
    // back commit
    List<HoodieRecord> records3 = tagLocation(index, writeRecords, hoodieTable).collect();
    assertEquals(0, records3.stream().filter(HoodieRecord::isCurrentLocationKnown).count());
    assertEquals(0, records3.stream().filter(record -> record.getCurrentLocation() != null).count());
}

Example 42 with SparkRDDWriteClient

use of org.apache.hudi.client.SparkRDDWriteClient in project hudi by apache.

the class TestSparkHoodieHBaseIndex method testSmallBatchSize.

@Test
public void testSmallBatchSize() throws Exception {
    final String newCommitTime = "001";
    final int numRecords = 10;
    List<HoodieRecord> records = dataGen.generateInserts(newCommitTime, numRecords);
    JavaRDD<HoodieRecord> writeRecords = jsc().parallelize(records, 1);
    // Load to memory
    HoodieWriteConfig config = getConfig(2);
    SparkHoodieHBaseIndex index = new SparkHoodieHBaseIndex(config);
    try (SparkRDDWriteClient writeClient = getHoodieWriteClient(config)) {
        metaClient = HoodieTableMetaClient.reload(metaClient);
        HoodieTable hoodieTable = HoodieSparkTable.create(config, context, metaClient);
        // Test tagLocation without any entries in index
        JavaRDD<HoodieRecord> records1 = tagLocation(index, writeRecords, hoodieTable);
        assertEquals(0, records1.filter(record -> record.isCurrentLocationKnown()).count());
        // Insert 200 records
        writeClient.startCommitWithTime(newCommitTime);
        JavaRDD<WriteStatus> writeStatues = writeClient.upsert(writeRecords, newCommitTime);
        assertNoWriteErrors(writeStatues.collect());
        // Now tagLocation for these records, hbaseIndex should not tag them since it was a failed
        // commit
        JavaRDD<HoodieRecord> records2 = tagLocation(index, writeRecords, hoodieTable);
        assertEquals(0, records2.filter(record -> record.isCurrentLocationKnown()).count());
        // Now commit this & update location of records inserted and validate no errors
        writeClient.commit(newCommitTime, writeStatues);
        // Now tagLocation for these records, hbaseIndex should tag them correctly
        metaClient = HoodieTableMetaClient.reload(metaClient);
        hoodieTable = HoodieSparkTable.create(config, context, metaClient);
        List<HoodieRecord> records3 = tagLocation(index, writeRecords, hoodieTable).collect();
        assertEquals(numRecords, records3.stream().filter(record -> record.isCurrentLocationKnown()).count());
        assertEquals(numRecords, records3.stream().map(record -> record.getKey().getRecordKey()).distinct().count());
        assertEquals(numRecords, records3.stream().filter(record -> (record.getCurrentLocation() != null && record.getCurrentLocation().getInstantTime().equals(newCommitTime))).distinct().count());
    }
}

Example 43 with SparkRDDWriteClient

use of org.apache.hudi.client.SparkRDDWriteClient in project hudi by apache.

the class TestSparkHoodieHBaseIndex method testTotalPutsBatching.

@Test
public void testTotalPutsBatching() throws Exception {
    HoodieWriteConfig config = getConfig();
    SparkHoodieHBaseIndex index = new SparkHoodieHBaseIndex(config);
    SparkRDDWriteClient writeClient = getHoodieWriteClient(config);
    // start a commit and generate test data
    String newCommitTime = writeClient.startCommit();
    List<HoodieRecord> records = dataGen.generateInserts(newCommitTime, 250);
    JavaRDD<HoodieRecord> writeRecords = jsc().parallelize(records, 1);
    metaClient = HoodieTableMetaClient.reload(metaClient);
    HoodieTable hoodieTable = HoodieSparkTable.create(config, context, metaClient);
    // Insert 200 records
    JavaRDD<WriteStatus> writeStatues = writeClient.upsert(writeRecords, newCommitTime);
    // commit this upsert
    writeClient.commit(newCommitTime, writeStatues);
    // Mock hbaseConnection and related entities
    Connection hbaseConnection = mock(Connection.class);
    HTable table = mock(HTable.class);
    when(hbaseConnection.getTable(TableName.valueOf(TABLE_NAME))).thenReturn(table);
    when(table.get((List<Get>) any())).thenReturn(new Result[0]);
    // only for test, set the hbaseConnection to mocked object
    index.setHbaseConnection(hbaseConnection);
    // Get all the files generated
    int numberOfDataFileIds = (int) writeStatues.map(status -> status.getFileId()).distinct().count();
    updateLocation(index, writeStatues, hoodieTable);
    // 3 batches should be executed given batchSize = 100 and <=numberOfDataFileIds getting updated,
    // so each fileId ideally gets updates
    verify(table, atMost(numberOfDataFileIds)).put((List<Put>) any());
}

Example 44 with SparkRDDWriteClient

use of org.apache.hudi.client.SparkRDDWriteClient in project hudi by apache.

the class BootstrapExecutor method execute.

/**
 * Executes Bootstrap.
 */
public void execute() throws IOException {
    initializeTable();
    SparkRDDWriteClient bootstrapClient = new SparkRDDWriteClient(new HoodieSparkEngineContext(jssc), bootstrapConfig);
    try {
        HashMap<String, String> checkpointCommitMetadata = new HashMap<>();
        checkpointCommitMetadata.put(HoodieDeltaStreamer.CHECKPOINT_KEY, cfg.checkpoint);
        if (cfg.checkpoint != null) {
            checkpointCommitMetadata.put(HoodieDeltaStreamer.CHECKPOINT_RESET_KEY, cfg.checkpoint);
        }
        bootstrapClient.bootstrap(Option.of(checkpointCommitMetadata));
        syncHive();
    } finally {
        bootstrapClient.close();
    }
}

Example 45 with SparkRDDWriteClient

use of org.apache.hudi.client.SparkRDDWriteClient in project hudi by apache.

the class HoodieCleaner method run.

public void run() {
    HoodieWriteConfig hoodieCfg = getHoodieClientConfig();
    SparkRDDWriteClient client = new SparkRDDWriteClient<>(new HoodieSparkEngineContext(jssc), hoodieCfg);
    client.clean();
}