Examples with S3SpillLocation com.amazonaws.athena.connector.lambda.domain.spill.S3SpillLocation used on opensource projects

Example 6 with S3SpillLocation

use of com.amazonaws.athena.connector.lambda.domain.spill.S3SpillLocation in project aws-athena-query-federation by awslabs.

the class RedisRecordHandlerTest method doReadRecordsZset.

@Test
public void doReadRecordsZset() throws Exception {
    // 4 keys per prefix
    when(mockSyncCommands.scan(any(ScanCursor.class), any(ScanArgs.class))).then((InvocationOnMock invocationOnMock) -> {
        ScanCursor cursor = (ScanCursor) invocationOnMock.getArguments()[0];
        if (cursor == null || cursor.getCursor().equals("0")) {
            List<String> result = new ArrayList<>();
            result.add(UUID.randomUUID().toString());
            result.add(UUID.randomUUID().toString());
            result.add(UUID.randomUUID().toString());
            MockKeyScanCursor<String> scanCursor = new MockKeyScanCursor<>();
            scanCursor.setCursor("1");
            scanCursor.setKeys(result);
            return scanCursor;
        } else {
            List<String> result = new ArrayList<>();
            result.add(UUID.randomUUID().toString());
            MockKeyScanCursor<String> scanCursor = new MockKeyScanCursor<>();
            scanCursor.setCursor("0");
            scanCursor.setKeys(result);
            scanCursor.setFinished(true);
            return scanCursor;
        }
    });
    // 4 rows per key
    when(mockSyncCommands.zscan(anyString(), any(ScanCursor.class))).then((InvocationOnMock invocationOnMock) -> {
        ScanCursor cursor = (ScanCursor) invocationOnMock.getArguments()[1];
        if (cursor == null || cursor.getCursor().equals("0")) {
            List<ScoredValue<String>> result = new ArrayList<>();
            result.add(ScoredValue.just(0.0D, "1"));
            result.add(ScoredValue.just(0.0D, "2"));
            result.add(ScoredValue.just(0.0D, "3"));
            MockScoredValueScanCursor<String> scanCursor = new MockScoredValueScanCursor<>();
            scanCursor.setCursor("1");
            scanCursor.setValues(result);
            return scanCursor;
        } else {
            List<ScoredValue<String>> result = new ArrayList<>();
            result.add(ScoredValue.just(0.0D, "4"));
            MockScoredValueScanCursor<String> scanCursor = new MockScoredValueScanCursor<>();
            scanCursor.setCursor("0");
            scanCursor.setValues(result);
            scanCursor.setFinished(true);
            return scanCursor;
        }
    });
    AtomicLong value = new AtomicLong(0);
    when(mockSyncCommands.get(anyString())).thenAnswer((InvocationOnMock invocationOnMock) -> String.valueOf(value.getAndIncrement()));
    S3SpillLocation splitLoc = S3SpillLocation.newBuilder().withBucket(UUID.randomUUID().toString()).withSplitId(UUID.randomUUID().toString()).withQueryId(UUID.randomUUID().toString()).withIsDirectory(true).build();
    Split split = Split.newBuilder(splitLoc, keyFactory.create()).add(REDIS_ENDPOINT_PROP, endpoint).add(KEY_TYPE, KeyType.PREFIX.getId()).add(KEY_PREFIX_TABLE_PROP, "key-*").add(VALUE_TYPE_TABLE_PROP, ValueType.ZSET.getId()).build();
    Schema schemaForRead = SchemaBuilder.newBuilder().addField("_key_", Types.MinorType.VARCHAR.getType()).addField("intcol", Types.MinorType.INT.getType()).build();
    Map<String, ValueSet> constraintsMap = new HashMap<>();
    constraintsMap.put("intcol", SortedRangeSet.copyOf(Types.MinorType.INT.getType(), ImmutableList.of(Range.greaterThan(allocator, Types.MinorType.INT.getType(), 1)), false));
    ReadRecordsRequest request = new ReadRecordsRequest(IDENTITY, DEFAULT_CATALOG, "queryId-" + System.currentTimeMillis(), TABLE_NAME, schemaForRead, split, new Constraints(constraintsMap), // 100GB don't expect this to spill
    100_000_000_000L, 100_000_000_000L);
    RecordResponse rawResponse = handler.doReadRecords(allocator, request);
    assertTrue(rawResponse instanceof ReadRecordsResponse);
    ReadRecordsResponse response = (ReadRecordsResponse) rawResponse;
    logger.info("doReadRecordsZset: rows[{}]", response.getRecordCount());
    logger.info("doReadRecordsZset: {}", BlockUtils.rowToString(response.getRecords(), 0));
    assertTrue(response.getRecords().getRowCount() == 12);
    FieldReader keyReader = response.getRecords().getFieldReader(KEY_COLUMN_NAME);
    keyReader.setPosition(0);
    assertNotNull(keyReader.readText());
    FieldReader intCol = response.getRecords().getFieldReader("intcol");
    intCol.setPosition(0);
    assertNotNull(intCol.readInteger());
}

Example 7 with S3SpillLocation

use of com.amazonaws.athena.connector.lambda.domain.spill.S3SpillLocation in project aws-athena-query-federation by awslabs.

the class RedisRecordHandlerTest method doReadRecordsLiteral.

@Test
public void doReadRecordsLiteral() throws Exception {
    // 4 keys per prefix
    when(mockSyncCommands.scan(any(ScanCursor.class), any(ScanArgs.class))).then((InvocationOnMock invocationOnMock) -> {
        ScanCursor cursor = (ScanCursor) invocationOnMock.getArguments()[0];
        if (cursor == null || cursor.getCursor().equals("0")) {
            List<String> result = new ArrayList<>();
            result.add(UUID.randomUUID().toString());
            result.add(UUID.randomUUID().toString());
            result.add(UUID.randomUUID().toString());
            MockKeyScanCursor<String> scanCursor = new MockKeyScanCursor<>();
            scanCursor.setCursor("1");
            scanCursor.setKeys(result);
            return scanCursor;
        } else {
            List<String> result = new ArrayList<>();
            result.add(UUID.randomUUID().toString());
            MockKeyScanCursor<String> scanCursor = new MockKeyScanCursor<>();
            scanCursor.setCursor("0");
            scanCursor.setKeys(result);
            scanCursor.setFinished(true);
            return scanCursor;
        }
    });
    AtomicLong value = new AtomicLong(0);
    when(mockSyncCommands.get(anyString())).thenAnswer((InvocationOnMock invocationOnMock) -> String.valueOf(value.getAndIncrement()));
    S3SpillLocation splitLoc = S3SpillLocation.newBuilder().withBucket(UUID.randomUUID().toString()).withSplitId(UUID.randomUUID().toString()).withQueryId(UUID.randomUUID().toString()).withIsDirectory(true).build();
    Split split = Split.newBuilder(splitLoc, keyFactory.create()).add(REDIS_ENDPOINT_PROP, endpoint).add(KEY_TYPE, KeyType.PREFIX.getId()).add(KEY_PREFIX_TABLE_PROP, "key-*").add(VALUE_TYPE_TABLE_PROP, ValueType.LITERAL.getId()).build();
    Schema schemaForRead = SchemaBuilder.newBuilder().addField("_key_", Types.MinorType.VARCHAR.getType()).addField("intcol", Types.MinorType.INT.getType()).build();
    Map<String, ValueSet> constraintsMap = new HashMap<>();
    constraintsMap.put("intcol", SortedRangeSet.copyOf(Types.MinorType.INT.getType(), ImmutableList.of(Range.greaterThan(allocator, Types.MinorType.INT.getType(), 1)), false));
    ReadRecordsRequest request = new ReadRecordsRequest(IDENTITY, DEFAULT_CATALOG, "queryId-" + System.currentTimeMillis(), TABLE_NAME, schemaForRead, split, new Constraints(constraintsMap), // 100GB don't expect this to spill
    100_000_000_000L, 100_000_000_000L);
    RecordResponse rawResponse = handler.doReadRecords(allocator, request);
    assertTrue(rawResponse instanceof ReadRecordsResponse);
    ReadRecordsResponse response = (ReadRecordsResponse) rawResponse;
    logger.info("doReadRecordsLiteral: rows[{}]", response.getRecordCount());
    logger.info("doReadRecordsLiteral: {}", BlockUtils.rowToString(response.getRecords(), 0));
    assertTrue(response.getRecords().getRowCount() == 2);
    FieldReader keyReader = response.getRecords().getFieldReader(KEY_COLUMN_NAME);
    keyReader.setPosition(0);
    assertNotNull(keyReader.readText().toString());
    FieldReader intCol = response.getRecords().getFieldReader("intcol");
    intCol.setPosition(0);
    assertNotNull(intCol.readInteger());
}

Example 8 with S3SpillLocation

use of com.amazonaws.athena.connector.lambda.domain.spill.S3SpillLocation in project aws-athena-query-federation by awslabs.

the class TPCDSRecordHandlerTest method doReadRecordsSpill.

@Test
public void doReadRecordsSpill() throws Exception {
    logger.info("doReadRecordsSpill: enter");
    Map<String, ValueSet> constraintsMap = new HashMap<>();
    constraintsMap.put("c_current_cdemo_sk", SortedRangeSet.of(Range.range(allocator, Types.MinorType.BIGINT.getType(), 100L, true, 100_000_000L, true)));
    ReadRecordsRequest request = new ReadRecordsRequest(identity, "catalog", "queryId-" + System.currentTimeMillis(), new TableName("tpcds1", table.getName()), schemaForRead, Split.newBuilder(S3SpillLocation.newBuilder().withBucket(UUID.randomUUID().toString()).withSplitId(UUID.randomUUID().toString()).withQueryId(UUID.randomUUID().toString()).withIsDirectory(true).build(), keyFactory.create()).add(SPLIT_NUMBER_FIELD, "0").add(SPLIT_TOTAL_NUMBER_FIELD, "10000").add(SPLIT_SCALE_FACTOR_FIELD, "1").build(), new Constraints(constraintsMap), // ~1.5MB so we should see some spill
    1_500_000L, 0);
    RecordResponse rawResponse = handler.doReadRecords(allocator, request);
    assertTrue(rawResponse instanceof RemoteReadRecordsResponse);
    try (RemoteReadRecordsResponse response = (RemoteReadRecordsResponse) rawResponse) {
        logger.info("doReadRecordsSpill: remoteBlocks[{}]", response.getRemoteBlocks().size());
        assertTrue(response.getNumberBlocks() > 1);
        int blockNum = 0;
        for (SpillLocation next : response.getRemoteBlocks()) {
            S3SpillLocation spillLocation = (S3SpillLocation) next;
            try (Block block = spillReader.read(spillLocation, response.getEncryptionKey(), response.getSchema())) {
                logger.info("doReadRecordsSpill: blockNum[{}] and recordCount[{}]", blockNum++, block.getRowCount());
                // assertTrue(++blockNum < response.getRemoteBlocks().size() && block.getRowCount() > 10_000);
                logger.info("doReadRecordsSpill: {}", BlockUtils.rowToString(block, 0));
                assertNotNull(BlockUtils.rowToString(block, 0));
            }
        }
    }
    logger.info("doReadRecordsSpill: exit");
}

Example 9 with S3SpillLocation

use of com.amazonaws.athena.connector.lambda.domain.spill.S3SpillLocation in project aws-athena-query-federation by awslabs.

the class TimestreamRecordHandlerTest method doReadRecordsNoSpill.

@Test
public void doReadRecordsNoSpill() throws Exception {
    int numRowsGenerated = 1_000;
    String expectedQuery = "SELECT measure_name, measure_value::double, az, time, hostname, region FROM \"my_schema\".\"my_table\" WHERE (\"az\" IN ('us-east-1a','us-east-1b'))";
    QueryResult mockResult = makeMockQueryResult(schemaForRead, numRowsGenerated);
    when(mockClient.query(any(QueryRequest.class))).thenAnswer((Answer<QueryResult>) invocationOnMock -> {
        QueryRequest request = (QueryRequest) invocationOnMock.getArguments()[0];
        assertEquals(expectedQuery, request.getQueryString().replace("\n", ""));
        return mockResult;
    });
    Map<String, ValueSet> constraintsMap = new HashMap<>();
    constraintsMap.put("az", EquatableValueSet.newBuilder(allocator, Types.MinorType.VARCHAR.getType(), true, true).add("us-east-1a").add("us-east-1b").build());
    S3SpillLocation splitLoc = S3SpillLocation.newBuilder().withBucket(UUID.randomUUID().toString()).withSplitId(UUID.randomUUID().toString()).withQueryId(UUID.randomUUID().toString()).withIsDirectory(true).build();
    Split.Builder splitBuilder = Split.newBuilder(splitLoc, keyFactory.create());
    ReadRecordsRequest request = new ReadRecordsRequest(IDENTITY, DEFAULT_CATALOG, "queryId-" + System.currentTimeMillis(), new TableName(DEFAULT_SCHEMA, TEST_TABLE), schemaForRead, splitBuilder.build(), new Constraints(constraintsMap), // 100GB don't expect this to spill
    100_000_000_000L, 100_000_000_000L);
    RecordResponse rawResponse = handler.doReadRecords(allocator, request);
    assertTrue(rawResponse instanceof ReadRecordsResponse);
    ReadRecordsResponse response = (ReadRecordsResponse) rawResponse;
    logger.info("doReadRecordsNoSpill: rows[{}]", response.getRecordCount());
    assertTrue(response.getRecords().getRowCount() > 0);
    // ensure we actually filtered something out
    assertTrue(response.getRecords().getRowCount() < numRowsGenerated);
    logger.info("doReadRecordsNoSpill: {}", BlockUtils.rowToString(response.getRecords(), 0));
}

Example 10 with S3SpillLocation

use of com.amazonaws.athena.connector.lambda.domain.spill.S3SpillLocation in project aws-athena-query-federation by awslabs.

the class TimestreamRecordHandlerTest method readRecordsTimeSeriesView.

@Test
public void readRecordsTimeSeriesView() throws Exception {
    logger.info("readRecordsTimeSeriesView - enter");
    Schema schemaForReadView = SchemaBuilder.newBuilder().addField("region", Types.MinorType.VARCHAR.getType()).addField("az", Types.MinorType.VARCHAR.getType()).addField("hostname", Types.MinorType.VARCHAR.getType()).addField(FieldBuilder.newBuilder("cpu_utilization", Types.MinorType.LIST.getType()).addField(FieldBuilder.newBuilder("cpu_utilization", Types.MinorType.STRUCT.getType()).addDateMilliField("time").addFloat8Field("measure_value::double").build()).build()).addMetadata(VIEW_METADATA_FIELD, "select az, hostname, region,  CREATE_TIME_SERIES(time, measure_value::double) as cpu_utilization from \"" + DEFAULT_SCHEMA + "\".\"" + TEST_TABLE + "\" WHERE measure_name = 'cpu_utilization' GROUP BY measure_name, az, hostname, region").build();
    String expectedQuery = "WITH t1 AS ( select az, hostname, region,  CREATE_TIME_SERIES(time, measure_value::double) as cpu_utilization from \"my_schema\".\"my_table\" WHERE measure_name = 'cpu_utilization' GROUP BY measure_name, az, hostname, region )  SELECT region, az, hostname, cpu_utilization FROM t1 WHERE (\"az\" IN ('us-east-1a','us-east-1b'))";
    QueryResult mockResult = makeMockQueryResult(schemaForReadView, 1_000);
    when(mockClient.query(any(QueryRequest.class))).thenAnswer((Answer<QueryResult>) invocationOnMock -> {
        QueryRequest request = (QueryRequest) invocationOnMock.getArguments()[0];
        assertEquals("actual: " + request.getQueryString(), expectedQuery, request.getQueryString().replace("\n", ""));
        return mockResult;
    });
    S3SpillLocation splitLoc = S3SpillLocation.newBuilder().withBucket(UUID.randomUUID().toString()).withSplitId(UUID.randomUUID().toString()).withQueryId(UUID.randomUUID().toString()).withIsDirectory(true).build();
    Split split = Split.newBuilder(splitLoc, null).build();
    Map<String, ValueSet> constraintsMap = new HashMap<>();
    constraintsMap.put("az", EquatableValueSet.newBuilder(allocator, Types.MinorType.VARCHAR.getType(), true, true).add("us-east-1a").add("us-east-1b").build());
    ReadRecordsRequest request = new ReadRecordsRequest(IDENTITY, "default", "queryId-" + System.currentTimeMillis(), new TableName(DEFAULT_SCHEMA, TEST_TABLE), schemaForReadView, split, new Constraints(constraintsMap), // 100GB don't expect this to spill
    100_000_000_000L, 100_000_000_000L);
    RecordResponse rawResponse = handler.doReadRecords(allocator, request);
    ReadRecordsResponse response = (ReadRecordsResponse) rawResponse;
    logger.info("readRecordsTimeSeriesView: rows[{}]", response.getRecordCount());
    for (int i = 0; i < response.getRecordCount() && i < 10; i++) {
        logger.info("readRecordsTimeSeriesView: {}", BlockUtils.rowToString(response.getRecords(), i));
    }
    logger.info("readRecordsTimeSeriesView - exit");
}