Java Code Examples for org.apache.flink.core.fs.RecoverableWriter#getResumeRecoverableSerializer()

The following examples show how to use org.apache.flink.core.fs.RecoverableWriter#getResumeRecoverableSerializer() . You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.
Example 1
Source File: BucketStateSerializerTest.java    From Flink-CEPplus with Apache License 2.0 6 votes vote down vote up
@Test
public void testSerializationEmpty() throws IOException {
	final File testFolder = tempFolder.newFolder();
	final FileSystem fs = FileSystem.get(testFolder.toURI());
	final RecoverableWriter writer = fs.createRecoverableWriter();

	final Path testBucket = new Path(testFolder.getPath(), "test");

	final BucketState<String> bucketState = new BucketState<>(
			"test", testBucket, Long.MAX_VALUE, null, new HashMap<>());

	final SimpleVersionedSerializer<BucketState<String>> serializer =
			new BucketStateSerializer<>(
					writer.getResumeRecoverableSerializer(),
					writer.getCommitRecoverableSerializer(),
					SimpleVersionedStringSerializer.INSTANCE
			);

	byte[] bytes = SimpleVersionedSerialization.writeVersionAndSerialize(serializer, bucketState);
	final BucketState<String> recoveredState =  SimpleVersionedSerialization.readVersionAndDeSerialize(serializer, bytes);

	Assert.assertEquals(testBucket, recoveredState.getBucketPath());
	Assert.assertNull(recoveredState.getInProgressResumableFile());
	Assert.assertTrue(recoveredState.getCommittableFilesPerCheckpoint().isEmpty());
}
 
Example 2
Source File: BucketStateSerializerTest.java    From flink with Apache License 2.0 6 votes vote down vote up
@Test
public void testSerializationEmpty() throws IOException {
	final File testFolder = tempFolder.newFolder();
	final FileSystem fs = FileSystem.get(testFolder.toURI());
	final RecoverableWriter writer = fs.createRecoverableWriter();

	final Path testBucket = new Path(testFolder.getPath(), "test");

	final BucketState<String> bucketState = new BucketState<>(
			"test", testBucket, Long.MAX_VALUE, null, new HashMap<>());

	final SimpleVersionedSerializer<BucketState<String>> serializer =
			new BucketStateSerializer<>(
					writer.getResumeRecoverableSerializer(),
					writer.getCommitRecoverableSerializer(),
					SimpleVersionedStringSerializer.INSTANCE
			);

	byte[] bytes = SimpleVersionedSerialization.writeVersionAndSerialize(serializer, bucketState);
	final BucketState<String> recoveredState =  SimpleVersionedSerialization.readVersionAndDeSerialize(serializer, bytes);

	Assert.assertEquals(testBucket, recoveredState.getBucketPath());
	Assert.assertNull(recoveredState.getInProgressResumableFile());
	Assert.assertTrue(recoveredState.getCommittableFilesPerCheckpoint().isEmpty());
}
 
Example 3
Source File: HadoopS3RecoverableWriterITCase.java    From Flink-CEPplus with Apache License 2.0 5 votes vote down vote up
private void testResumeAfterMultiplePersist(
		final String persistName,
		final String expectedFinalContents,
		final String firstItemToWrite,
		final String secondItemToWrite,
		final String thirdItemToWrite) throws Exception {

	final Path path = new Path(basePathForTest, "part-0");
	final RecoverableWriter initWriter = getRecoverableWriter();

	final Map<String, RecoverableWriter.ResumeRecoverable> recoverables = new HashMap<>(4);
	try (final RecoverableFsDataOutputStream stream = initWriter.open(path)) {
		recoverables.put(INIT_EMPTY_PERSIST, stream.persist());

		stream.write(bytesOf(firstItemToWrite));

		recoverables.put(INTERM_WITH_STATE_PERSIST, stream.persist());
		recoverables.put(INTERM_WITH_NO_ADDITIONAL_STATE_PERSIST, stream.persist());

		// and write some more data
		stream.write(bytesOf(secondItemToWrite));

		recoverables.put(FINAL_WITH_EXTRA_STATE, stream.persist());
	}

	final SimpleVersionedSerializer<RecoverableWriter.ResumeRecoverable> serializer = initWriter.getResumeRecoverableSerializer();
	final byte[] serializedRecoverable = serializer.serialize(recoverables.get(persistName));

	// get a new serializer from a new writer to make sure that no pre-initialized state leaks in.
	final RecoverableWriter newWriter = getRecoverableWriter();
	final SimpleVersionedSerializer<RecoverableWriter.ResumeRecoverable> deserializer = newWriter.getResumeRecoverableSerializer();
	final RecoverableWriter.ResumeRecoverable recoveredRecoverable =
			deserializer.deserialize(serializer.getVersion(), serializedRecoverable);

	final RecoverableFsDataOutputStream recoveredStream = newWriter.recover(recoveredRecoverable);
	recoveredStream.write(bytesOf(thirdItemToWrite));
	recoveredStream.closeForCommit().commit();

	Assert.assertEquals(expectedFinalContents, getContentsOfFile(path));
}
 
Example 4
Source File: BucketStateSerializerTest.java    From Flink-CEPplus with Apache License 2.0 5 votes vote down vote up
@Test
public void testSerializationOnlyInProgress() throws IOException {
	final File testFolder = tempFolder.newFolder();
	final FileSystem fs = FileSystem.get(testFolder.toURI());

	final Path testBucket = new Path(testFolder.getPath(), "test");

	final RecoverableWriter writer = fs.createRecoverableWriter();
	final RecoverableFsDataOutputStream stream = writer.open(testBucket);
	stream.write(IN_PROGRESS_CONTENT.getBytes(Charset.forName("UTF-8")));

	final RecoverableWriter.ResumeRecoverable current = stream.persist();

	final BucketState<String> bucketState = new BucketState<>(
			"test", testBucket, Long.MAX_VALUE, current, new HashMap<>());

	final SimpleVersionedSerializer<BucketState<String>> serializer =
			new BucketStateSerializer<>(
					writer.getResumeRecoverableSerializer(),
					writer.getCommitRecoverableSerializer(),
					SimpleVersionedStringSerializer.INSTANCE
			);

	final byte[] bytes = SimpleVersionedSerialization.writeVersionAndSerialize(serializer, bucketState);

	// to simulate that everything is over for file.
	stream.close();

	final BucketState<String> recoveredState =  SimpleVersionedSerialization.readVersionAndDeSerialize(serializer, bytes);

	Assert.assertEquals(testBucket, recoveredState.getBucketPath());

	FileStatus[] statuses = fs.listStatus(testBucket.getParent());
	Assert.assertEquals(1L, statuses.length);
	Assert.assertTrue(
			statuses[0].getPath().getPath().startsWith(
					(new Path(testBucket.getParent(), ".test.inprogress")).toString())
	);
}
 
Example 5
Source File: HadoopS3RecoverableWriterITCase.java    From flink with Apache License 2.0 5 votes vote down vote up
private void testResumeAfterMultiplePersist(
		final String persistName,
		final String expectedFinalContents,
		final String firstItemToWrite,
		final String secondItemToWrite,
		final String thirdItemToWrite) throws Exception {

	final Path path = new Path(basePathForTest, "part-0");
	final RecoverableWriter initWriter = getRecoverableWriter();

	final Map<String, RecoverableWriter.ResumeRecoverable> recoverables = new HashMap<>(4);
	try (final RecoverableFsDataOutputStream stream = initWriter.open(path)) {
		recoverables.put(INIT_EMPTY_PERSIST, stream.persist());

		stream.write(bytesOf(firstItemToWrite));

		recoverables.put(INTERM_WITH_STATE_PERSIST, stream.persist());
		recoverables.put(INTERM_WITH_NO_ADDITIONAL_STATE_PERSIST, stream.persist());

		// and write some more data
		stream.write(bytesOf(secondItemToWrite));

		recoverables.put(FINAL_WITH_EXTRA_STATE, stream.persist());
	}

	final SimpleVersionedSerializer<RecoverableWriter.ResumeRecoverable> serializer = initWriter.getResumeRecoverableSerializer();
	final byte[] serializedRecoverable = serializer.serialize(recoverables.get(persistName));

	// get a new serializer from a new writer to make sure that no pre-initialized state leaks in.
	final RecoverableWriter newWriter = getRecoverableWriter();
	final SimpleVersionedSerializer<RecoverableWriter.ResumeRecoverable> deserializer = newWriter.getResumeRecoverableSerializer();
	final RecoverableWriter.ResumeRecoverable recoveredRecoverable =
			deserializer.deserialize(serializer.getVersion(), serializedRecoverable);

	final RecoverableFsDataOutputStream recoveredStream = newWriter.recover(recoveredRecoverable);
	recoveredStream.write(bytesOf(thirdItemToWrite));
	recoveredStream.closeForCommit().commit();

	Assert.assertEquals(expectedFinalContents, getContentsOfFile(path));
}
 
Example 6
Source File: BucketStateSerializerTest.java    From flink with Apache License 2.0 5 votes vote down vote up
@Test
public void testSerializationOnlyInProgress() throws IOException {
	final File testFolder = tempFolder.newFolder();
	final FileSystem fs = FileSystem.get(testFolder.toURI());

	final Path testBucket = new Path(testFolder.getPath(), "test");

	final RecoverableWriter writer = fs.createRecoverableWriter();
	final RecoverableFsDataOutputStream stream = writer.open(testBucket);
	stream.write(IN_PROGRESS_CONTENT.getBytes(Charset.forName("UTF-8")));

	final RecoverableWriter.ResumeRecoverable current = stream.persist();

	final BucketState<String> bucketState = new BucketState<>(
			"test", testBucket, Long.MAX_VALUE, current, new HashMap<>());

	final SimpleVersionedSerializer<BucketState<String>> serializer =
			new BucketStateSerializer<>(
					writer.getResumeRecoverableSerializer(),
					writer.getCommitRecoverableSerializer(),
					SimpleVersionedStringSerializer.INSTANCE
			);

	final byte[] bytes = SimpleVersionedSerialization.writeVersionAndSerialize(serializer, bucketState);

	// to simulate that everything is over for file.
	stream.close();

	final BucketState<String> recoveredState =  SimpleVersionedSerialization.readVersionAndDeSerialize(serializer, bytes);

	Assert.assertEquals(testBucket, recoveredState.getBucketPath());

	FileStatus[] statuses = fs.listStatus(testBucket.getParent());
	Assert.assertEquals(1L, statuses.length);
	Assert.assertTrue(
			statuses[0].getPath().getPath().startsWith(
					(new Path(testBucket.getParent(), ".test.inprogress")).getPath())
	);
}
 
Example 7
Source File: HadoopS3RecoverableWriterITCase.java    From flink with Apache License 2.0 5 votes vote down vote up
private void testResumeAfterMultiplePersist(
		final String persistName,
		final String expectedFinalContents,
		final String firstItemToWrite,
		final String secondItemToWrite,
		final String thirdItemToWrite) throws Exception {

	final Path path = new Path(basePathForTest, "part-0");
	final RecoverableWriter initWriter = getRecoverableWriter();

	final Map<String, RecoverableWriter.ResumeRecoverable> recoverables = new HashMap<>(4);
	try (final RecoverableFsDataOutputStream stream = initWriter.open(path)) {
		recoverables.put(INIT_EMPTY_PERSIST, stream.persist());

		stream.write(bytesOf(firstItemToWrite));

		recoverables.put(INTERM_WITH_STATE_PERSIST, stream.persist());
		recoverables.put(INTERM_WITH_NO_ADDITIONAL_STATE_PERSIST, stream.persist());

		// and write some more data
		stream.write(bytesOf(secondItemToWrite));

		recoverables.put(FINAL_WITH_EXTRA_STATE, stream.persist());
	}

	final SimpleVersionedSerializer<RecoverableWriter.ResumeRecoverable> serializer = initWriter.getResumeRecoverableSerializer();
	final byte[] serializedRecoverable = serializer.serialize(recoverables.get(persistName));

	// get a new serializer from a new writer to make sure that no pre-initialized state leaks in.
	final RecoverableWriter newWriter = getRecoverableWriter();
	final SimpleVersionedSerializer<RecoverableWriter.ResumeRecoverable> deserializer = newWriter.getResumeRecoverableSerializer();
	final RecoverableWriter.ResumeRecoverable recoveredRecoverable =
			deserializer.deserialize(serializer.getVersion(), serializedRecoverable);

	final RecoverableFsDataOutputStream recoveredStream = newWriter.recover(recoveredRecoverable);
	recoveredStream.write(bytesOf(thirdItemToWrite));
	recoveredStream.closeForCommit().commit();

	Assert.assertEquals(expectedFinalContents, getContentsOfFile(path));
}