use of io.druid.segment.loading.DataSegmentPusher in project druid by druid-io.
the class SameIntervalMergeTaskTest method runTask.
private List<DataSegment> runTask(final SameIntervalMergeTask mergeTask, final String version) throws Exception {
boolean isReady = mergeTask.isReady(new TaskActionClient() {
@Override
public <RetType> RetType submit(TaskAction<RetType> taskAction) throws IOException {
if (taskAction instanceof LockTryAcquireAction) {
// the lock of this interval is required
Assert.assertEquals(mergeTask.getInterval(), ((LockTryAcquireAction) taskAction).getInterval());
isRedayCountDown.countDown();
taskLock = new TaskLock(mergeTask.getGroupId(), mergeTask.getDataSource(), mergeTask.getInterval(), version);
return (RetType) taskLock;
}
return null;
}
});
// ensure LockTryAcquireAction is submitted
Assert.assertTrue(isReady);
final List<DataSegment> segments = Lists.newArrayList();
mergeTask.run(new TaskToolbox(null, null, new TaskActionClient() {
@Override
public <RetType> RetType submit(TaskAction<RetType> taskAction) throws IOException {
if (taskAction instanceof LockListAction) {
Assert.assertNotNull("taskLock should be acquired before list", taskLock);
return (RetType) Arrays.asList(taskLock);
}
if (taskAction instanceof SegmentListUsedAction) {
List<DataSegment> segments = ImmutableList.of(DataSegment.builder().dataSource(mergeTask.getDataSource()).interval(new Interval("2010-01-01/PT1H")).version("oldVersion").shardSpec(new LinearShardSpec(0)).build(), DataSegment.builder().dataSource(mergeTask.getDataSource()).interval(new Interval("2010-01-01/PT1H")).version("oldVersion").shardSpec(new LinearShardSpec(0)).build(), DataSegment.builder().dataSource(mergeTask.getDataSource()).interval(new Interval("2010-01-01/PT2H")).version("oldVersion").shardSpec(new LinearShardSpec(0)).build());
return (RetType) segments;
}
if (taskAction instanceof SegmentInsertAction) {
publishCountDown.countDown();
return null;
}
return null;
}
}, new NoopServiceEmitter(), new DataSegmentPusher() {
@Deprecated
@Override
public String getPathForHadoop(String dataSource) {
return getPathForHadoop();
}
@Override
public String getPathForHadoop() {
return null;
}
@Override
public DataSegment push(File file, DataSegment segment) throws IOException {
// the merged segment is pushed to storage
segments.add(segment);
return segment;
}
}, null, null, null, null, null, null, null, null, new SegmentLoader() {
@Override
public boolean isSegmentLoaded(DataSegment segment) throws SegmentLoadingException {
return false;
}
@Override
public Segment getSegment(DataSegment segment) throws SegmentLoadingException {
return null;
}
@Override
public File getSegmentFiles(DataSegment segment) throws SegmentLoadingException {
// dummy file to represent the downloaded segment's dir
return new File("" + segment.getShardSpec().getPartitionNum());
}
@Override
public void cleanup(DataSegment segment) throws SegmentLoadingException {
}
}, jsonMapper, temporaryFolder.newFolder(), EasyMock.createMock(IndexMerger.class), indexIO, null, null, EasyMock.createMock(IndexMergerV9.class)));
return segments;
}
use of io.druid.segment.loading.DataSegmentPusher in project druid by druid-io.
the class IndexTaskTest method runTask.
private final List<DataSegment> runTask(final IndexTask indexTask) throws Exception {
final List<DataSegment> segments = Lists.newArrayList();
indexTask.run(new TaskToolbox(null, null, new TaskActionClient() {
@Override
public <RetType> RetType submit(TaskAction<RetType> taskAction) throws IOException {
if (taskAction instanceof LockListAction) {
return (RetType) Arrays.asList(new TaskLock("", "", null, new DateTime().toString()));
}
if (taskAction instanceof LockAcquireAction) {
return (RetType) new TaskLock("groupId", "test", ((LockAcquireAction) taskAction).getInterval(), new DateTime().toString());
}
if (taskAction instanceof SegmentTransactionalInsertAction) {
return (RetType) new SegmentPublishResult(((SegmentTransactionalInsertAction) taskAction).getSegments(), true);
}
if (taskAction instanceof SegmentAllocateAction) {
SegmentAllocateAction action = (SegmentAllocateAction) taskAction;
Interval interval = action.getPreferredSegmentGranularity().bucket(action.getTimestamp());
ShardSpec shardSpec = new NumberedShardSpec(segmentAllocatePartitionCounter++, 0);
return (RetType) new SegmentIdentifier(action.getDataSource(), interval, "latestVersion", shardSpec);
}
return null;
}
}, null, new DataSegmentPusher() {
@Deprecated
@Override
public String getPathForHadoop(String dataSource) {
return getPathForHadoop();
}
@Override
public String getPathForHadoop() {
return null;
}
@Override
public DataSegment push(File file, DataSegment segment) throws IOException {
segments.add(segment);
return segment;
}
}, null, null, null, null, null, null, null, null, null, jsonMapper, temporaryFolder.newFolder(), indexMerger, indexIO, null, null, indexMergerV9));
Collections.sort(segments);
return segments;
}
use of io.druid.segment.loading.DataSegmentPusher in project druid by druid-io.
the class RealtimePlumberSchoolTest method setUp.
@Before
public void setUp() throws Exception {
tmpDir = Files.createTempDir();
ObjectMapper jsonMapper = new DefaultObjectMapper();
schema = new DataSchema("test", jsonMapper.convertValue(new StringInputRowParser(new JSONParseSpec(new TimestampSpec("timestamp", "auto", null), new DimensionsSpec(null, null, null), null, null), null), Map.class), new AggregatorFactory[] { new CountAggregatorFactory("rows") }, new UniformGranularitySpec(Granularities.HOUR, Granularities.NONE, null), jsonMapper);
schema2 = new DataSchema("test", jsonMapper.convertValue(new StringInputRowParser(new JSONParseSpec(new TimestampSpec("timestamp", "auto", null), new DimensionsSpec(null, null, null), null, null), null), Map.class), new AggregatorFactory[] { new CountAggregatorFactory("rows") }, new UniformGranularitySpec(Granularities.YEAR, Granularities.NONE, null), jsonMapper);
announcer = EasyMock.createMock(DataSegmentAnnouncer.class);
announcer.announceSegment(EasyMock.<DataSegment>anyObject());
EasyMock.expectLastCall().anyTimes();
segmentPublisher = EasyMock.createNiceMock(SegmentPublisher.class);
dataSegmentPusher = EasyMock.createNiceMock(DataSegmentPusher.class);
handoffNotifierFactory = EasyMock.createNiceMock(SegmentHandoffNotifierFactory.class);
handoffNotifier = EasyMock.createNiceMock(SegmentHandoffNotifier.class);
EasyMock.expect(handoffNotifierFactory.createSegmentHandoffNotifier(EasyMock.anyString())).andReturn(handoffNotifier).anyTimes();
EasyMock.expect(handoffNotifier.registerSegmentHandoffCallback(EasyMock.<SegmentDescriptor>anyObject(), EasyMock.<Executor>anyObject(), EasyMock.<Runnable>anyObject())).andReturn(true).anyTimes();
emitter = EasyMock.createMock(ServiceEmitter.class);
EasyMock.replay(announcer, segmentPublisher, dataSegmentPusher, handoffNotifierFactory, handoffNotifier, emitter);
tuningConfig = new RealtimeTuningConfig(1, null, null, null, new IntervalStartVersioningPolicy(), rejectionPolicy, null, null, null, buildV9Directly, 0, 0, false, null);
realtimePlumberSchool = new RealtimePlumberSchool(emitter, new DefaultQueryRunnerFactoryConglomerate(Maps.<Class<? extends Query>, QueryRunnerFactory>newHashMap()), dataSegmentPusher, announcer, segmentPublisher, handoffNotifierFactory, MoreExecutors.sameThreadExecutor(), TestHelper.getTestIndexMerger(), TestHelper.getTestIndexMergerV9(), TestHelper.getTestIndexIO(), MapCache.create(0), FireDepartmentTest.NO_CACHE_CONFIG, TestHelper.getObjectMapper());
metrics = new FireDepartmentMetrics();
plumber = (RealtimePlumber) realtimePlumberSchool.findPlumber(schema, tuningConfig, metrics);
}
use of io.druid.segment.loading.DataSegmentPusher in project druid by druid-io.
the class KafkaIndexTaskTest method makeToolboxFactory.
private void makeToolboxFactory() throws IOException {
directory = tempFolder.newFolder();
final TestUtils testUtils = new TestUtils();
final ObjectMapper objectMapper = testUtils.getTestObjectMapper();
for (Module module : new KafkaIndexTaskModule().getJacksonModules()) {
objectMapper.registerModule(module);
}
final TaskConfig taskConfig = new TaskConfig(new File(directory, "taskBaseDir").getPath(), null, null, 50000, null, false, null, null);
final TestDerbyConnector derbyConnector = derby.getConnector();
derbyConnector.createDataSourceTable();
derbyConnector.createPendingSegmentsTable();
derbyConnector.createSegmentTable();
derbyConnector.createRulesTable();
derbyConnector.createConfigTable();
derbyConnector.createTaskTables();
derbyConnector.createAuditTable();
taskStorage = new MetadataTaskStorage(derbyConnector, new TaskStorageConfig(null), new SQLMetadataStorageActionHandlerFactory(derbyConnector, derby.metadataTablesConfigSupplier().get(), objectMapper));
metadataStorageCoordinator = new IndexerSQLMetadataStorageCoordinator(testUtils.getTestObjectMapper(), derby.metadataTablesConfigSupplier().get(), derbyConnector);
taskLockbox = new TaskLockbox(taskStorage);
final TaskActionToolbox taskActionToolbox = new TaskActionToolbox(taskLockbox, metadataStorageCoordinator, emitter, new SupervisorManager(null));
final TaskActionClientFactory taskActionClientFactory = new LocalTaskActionClientFactory(taskStorage, taskActionToolbox);
final SegmentHandoffNotifierFactory handoffNotifierFactory = new SegmentHandoffNotifierFactory() {
@Override
public SegmentHandoffNotifier createSegmentHandoffNotifier(String dataSource) {
return new SegmentHandoffNotifier() {
@Override
public boolean registerSegmentHandoffCallback(SegmentDescriptor descriptor, Executor exec, Runnable handOffRunnable) {
if (doHandoff) {
// Simulate immediate handoff
exec.execute(handOffRunnable);
}
return true;
}
@Override
public void start() {
//Noop
}
@Override
public void close() {
//Noop
}
};
}
};
final LocalDataSegmentPusherConfig dataSegmentPusherConfig = new LocalDataSegmentPusherConfig();
dataSegmentPusherConfig.storageDirectory = getSegmentDirectory();
final DataSegmentPusher dataSegmentPusher = new LocalDataSegmentPusher(dataSegmentPusherConfig, objectMapper);
toolboxFactory = new TaskToolboxFactory(taskConfig, taskActionClientFactory, emitter, dataSegmentPusher, new TestDataSegmentKiller(), // DataSegmentMover
null, // DataSegmentArchiver
null, new TestDataSegmentAnnouncer(), handoffNotifierFactory, makeTimeseriesOnlyConglomerate(), // queryExecutorService
MoreExecutors.sameThreadExecutor(), EasyMock.createMock(MonitorScheduler.class), new SegmentLoaderFactory(new SegmentLoaderLocalCacheManager(null, new SegmentLoaderConfig() {
@Override
public List<StorageLocationConfig> getLocations() {
return Lists.newArrayList();
}
}, testUtils.getTestObjectMapper())), testUtils.getTestObjectMapper(), testUtils.getTestIndexMerger(), testUtils.getTestIndexIO(), MapCache.create(1024), new CacheConfig(), testUtils.getTestIndexMergerV9());
}
use of io.druid.segment.loading.DataSegmentPusher in project druid by druid-io.
the class IngestSegmentFirehoseFactoryTest method constructorFeeder.
@Parameterized.Parameters(name = "{1}")
public static Collection<Object[]> constructorFeeder() throws IOException {
final IndexSpec indexSpec = new IndexSpec();
final HeapMemoryTaskStorage ts = new HeapMemoryTaskStorage(new TaskStorageConfig(null) {
});
final IncrementalIndexSchema schema = new IncrementalIndexSchema.Builder().withQueryGranularity(Granularities.NONE).withMinTimestamp(JodaUtils.MIN_INSTANT).withDimensionsSpec(ROW_PARSER).withMetrics(new AggregatorFactory[] { new LongSumAggregatorFactory(METRIC_LONG_NAME, DIM_LONG_NAME), new DoubleSumAggregatorFactory(METRIC_FLOAT_NAME, DIM_FLOAT_NAME) }).build();
final OnheapIncrementalIndex index = new OnheapIncrementalIndex(schema, true, MAX_ROWS * MAX_SHARD_NUMBER);
for (Integer i = 0; i < MAX_ROWS; ++i) {
index.add(ROW_PARSER.parse(buildRow(i.longValue())));
}
if (!persistDir.mkdirs() && !persistDir.exists()) {
throw new IOException(String.format("Could not create directory at [%s]", persistDir.getAbsolutePath()));
}
INDEX_MERGER.persist(index, persistDir, indexSpec);
final TaskLockbox tl = new TaskLockbox(ts);
final IndexerSQLMetadataStorageCoordinator mdc = new IndexerSQLMetadataStorageCoordinator(null, null, null) {
private final Set<DataSegment> published = Sets.newHashSet();
private final Set<DataSegment> nuked = Sets.newHashSet();
@Override
public List<DataSegment> getUsedSegmentsForInterval(String dataSource, Interval interval) throws IOException {
return ImmutableList.copyOf(segmentSet);
}
@Override
public List<DataSegment> getUsedSegmentsForIntervals(String dataSource, List<Interval> interval) throws IOException {
return ImmutableList.copyOf(segmentSet);
}
@Override
public List<DataSegment> getUnusedSegmentsForInterval(String dataSource, Interval interval) {
return ImmutableList.of();
}
@Override
public Set<DataSegment> announceHistoricalSegments(Set<DataSegment> segments) {
Set<DataSegment> added = Sets.newHashSet();
for (final DataSegment segment : segments) {
if (published.add(segment)) {
added.add(segment);
}
}
return ImmutableSet.copyOf(added);
}
@Override
public void deleteSegments(Set<DataSegment> segments) {
nuked.addAll(segments);
}
};
final LocalTaskActionClientFactory tac = new LocalTaskActionClientFactory(ts, new TaskActionToolbox(tl, mdc, newMockEmitter(), EasyMock.createMock(SupervisorManager.class)));
SegmentHandoffNotifierFactory notifierFactory = EasyMock.createNiceMock(SegmentHandoffNotifierFactory.class);
EasyMock.replay(notifierFactory);
final TaskToolboxFactory taskToolboxFactory = new TaskToolboxFactory(new TaskConfig(tmpDir.getAbsolutePath(), null, null, 50000, null, false, null, null), tac, newMockEmitter(), new DataSegmentPusher() {
@Deprecated
@Override
public String getPathForHadoop(String dataSource) {
return getPathForHadoop();
}
@Override
public String getPathForHadoop() {
throw new UnsupportedOperationException();
}
@Override
public DataSegment push(File file, DataSegment segment) throws IOException {
return segment;
}
}, new DataSegmentKiller() {
@Override
public void kill(DataSegment segments) throws SegmentLoadingException {
}
@Override
public void killAll() throws IOException {
throw new UnsupportedOperationException("not implemented");
}
}, new DataSegmentMover() {
@Override
public DataSegment move(DataSegment dataSegment, Map<String, Object> targetLoadSpec) throws SegmentLoadingException {
return dataSegment;
}
}, new DataSegmentArchiver() {
@Override
public DataSegment archive(DataSegment segment) throws SegmentLoadingException {
return segment;
}
@Override
public DataSegment restore(DataSegment segment) throws SegmentLoadingException {
return segment;
}
}, // segment announcer
null, notifierFactory, // query runner factory conglomerate corporation unionized collective
null, // query executor service
null, // monitor scheduler
null, new SegmentLoaderFactory(new SegmentLoaderLocalCacheManager(null, new SegmentLoaderConfig() {
@Override
public List<StorageLocationConfig> getLocations() {
return Lists.newArrayList();
}
}, MAPPER)), MAPPER, INDEX_MERGER, INDEX_IO, null, null, INDEX_MERGER_V9);
Collection<Object[]> values = new LinkedList<>();
for (InputRowParser parser : Arrays.<InputRowParser>asList(ROW_PARSER, new MapInputRowParser(new JSONParseSpec(new TimestampSpec(TIME_COLUMN, "auto", null), new DimensionsSpec(DimensionsSpec.getDefaultSchemas(ImmutableList.<String>of()), ImmutableList.of(DIM_FLOAT_NAME, DIM_LONG_NAME), ImmutableList.<SpatialDimensionSchema>of()), null, null)))) {
for (List<String> dim_names : Arrays.<List<String>>asList(null, ImmutableList.of(DIM_NAME))) {
for (List<String> metric_names : Arrays.<List<String>>asList(null, ImmutableList.of(METRIC_LONG_NAME, METRIC_FLOAT_NAME))) {
values.add(new Object[] { new IngestSegmentFirehoseFactory(DATA_SOURCE_NAME, FOREVER, new SelectorDimFilter(DIM_NAME, DIM_VALUE, null), dim_names, metric_names, Guice.createInjector(new Module() {
@Override
public void configure(Binder binder) {
binder.bind(TaskToolboxFactory.class).toInstance(taskToolboxFactory);
}
}), INDEX_IO), String.format("DimNames[%s]MetricNames[%s]ParserDimNames[%s]", dim_names == null ? "null" : "dims", metric_names == null ? "null" : "metrics", parser == ROW_PARSER ? "dims" : "null"), parser });
}
}
}
return values;
}
Aggregations