@Test
public void testReadingExistingVersion1HFile() throws IOException {
URL url = TestHFileReaderV1.class.getResource("8e8ab58dcf39412da19833fcd8f687ac");
Path existingHFilePath = new Path(url.getPath());
HFile.Reader reader = HFile.createReader(fs, existingHFilePath, new CacheConfig(conf));
reader.loadFileInfo();
FixedFileTrailer trailer = reader.getTrailer();
assertEquals(N, reader.getEntries());
assertEquals(N, trailer.getEntryCount());
assertEquals(1, trailer.getMajorVersion());
assertEquals(Compression.Algorithm.GZ, trailer.getCompressionCodec());
for (boolean pread : new boolean[] {false, true}) {
int totalDataSize = 0;
int n = 0;
HFileScanner scanner = reader.getScanner(false, pread);
assertTrue(scanner.seekTo());
do {
totalDataSize +=
scanner.getKey().limit() + scanner.getValue().limit() + Bytes.SIZEOF_INT * 2;
++n;
} while (scanner.next());
// Add magic record sizes, one per data block.
totalDataSize += 8 * trailer.getDataIndexCount();
assertEquals(N, n);
assertEquals(trailer.getTotalUncompressedBytes(), totalDataSize);
}
reader.close();
}
@Test(timeout = 6000000)
public void testHFileEncryption() throws Exception {
// Create 1000 random test KVs
RedundantKVGenerator generator = new RedundantKVGenerator();
List<KeyValue> testKvs = generator.generateTestKeyValues(1000);
// Iterate through data block encoding and compression combinations
Configuration conf = TEST_UTIL.getConfiguration();
CacheConfig cacheConf = new CacheConfig(conf);
for (DataBlockEncoding encoding : DataBlockEncoding.values()) {
for (Compression.Algorithm compression : TestHFileBlock.COMPRESSION_ALGORITHMS) {
HFileContext fileContext =
new HFileContextBuilder()
.withBlockSize(4096) // small blocks
.withEncryptionContext(cryptoContext)
.withCompression(compression)
.withDataBlockEncoding(encoding)
.build();
// write a new test HFile
LOG.info("Writing with " + fileContext);
Path path = new Path(TEST_UTIL.getDataTestDir(), UUID.randomUUID().toString() + ".hfile");
FSDataOutputStream out = fs.create(path);
HFile.Writer writer =
HFile.getWriterFactory(conf, cacheConf)
.withOutputStream(out)
.withFileContext(fileContext)
.create();
try {
for (KeyValue kv : testKvs) {
writer.append(kv);
}
} finally {
writer.close();
out.close();
}
// read it back in
LOG.info("Reading with " + fileContext);
int i = 0;
HFileScanner scanner = null;
HFile.Reader reader = HFile.createReader(fs, path, cacheConf, conf);
try {
reader.loadFileInfo();
FixedFileTrailer trailer = reader.getTrailer();
assertNotNull(trailer.getEncryptionKey());
scanner = reader.getScanner(false, false);
assertTrue("Initial seekTo failed", scanner.seekTo());
do {
Cell kv = scanner.getCell();
assertTrue(
"Read back an unexpected or invalid KV",
testKvs.contains(KeyValueUtil.ensureKeyValue(kv)));
i++;
} while (scanner.next());
} finally {
reader.close();
scanner.close();
}
assertEquals("Did not read back as many KVs as written", i, testKvs.size());
// Test random seeks with pread
LOG.info("Random seeking with " + fileContext);
reader = HFile.createReader(fs, path, cacheConf, conf);
try {
scanner = reader.getScanner(false, true);
assertTrue("Initial seekTo failed", scanner.seekTo());
for (i = 0; i < 100; i++) {
KeyValue kv = testKvs.get(RNG.nextInt(testKvs.size()));
assertEquals("Unable to find KV as expected: " + kv, scanner.seekTo(kv), 0);
}
} finally {
scanner.close();
reader.close();
}
}
}
}