/**
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.hadoop.raid;
import java.io.File;
import java.io.FileWriter;
import java.io.IOException;
import java.lang.management.ManagementFactory;
import java.net.URI;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.HashSet;
import java.util.LinkedList;
import java.util.List;
import java.util.Map;
import java.util.Random;
import java.util.Set;
import javax.management.MBeanServer;
import javax.management.ObjectName;
import junit.framework.TestCase;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileStatus;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.hdfs.DFSUtil;
import org.apache.hadoop.hdfs.DistributedFileSystem;
import org.apache.hadoop.hdfs.MiniDFSCluster;
import org.apache.hadoop.hdfs.RaidDFSUtil;
import org.apache.hadoop.hdfs.TestDatanodeBlockScanner;
import org.apache.hadoop.hdfs.TestRaidDfs;
import org.apache.hadoop.hdfs.protocol.LocatedBlock;
import org.apache.hadoop.hdfs.protocol.Block;
import org.apache.hadoop.hdfs.protocol.LocatedBlocks;
import org.apache.hadoop.hdfs.util.InjectionEvent;
import org.apache.hadoop.mapred.MiniMRCluster;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.JobID;
import org.apache.hadoop.metrics.util.MetricsTimeVaryingLong;
import org.apache.hadoop.raid.DistBlockIntegrityMonitor.Priority;
import org.apache.hadoop.raid.DistBlockIntegrityMonitor.Worker.LostFileInfo;
import org.apache.hadoop.raid.LogUtils.LOGRESULTS;
import org.apache.hadoop.raid.LogUtils.LOGTYPES;
import org.apache.hadoop.raid.RaidHistogram.Point;
import org.apache.hadoop.util.InjectionEventI;
import org.apache.hadoop.util.InjectionHandler;
import org.apache.hadoop.util.StringUtils;
import org.junit.Test;
public class TestBlockFixer extends TestCase {
final static Log LOG = LogFactory.getLog(
"org.apache.hadoop.raid.TestBlockFixer");
final static String TEST_DIR = new File(System.getProperty("test.build.data",
"build/contrib/raid/test/data")).getAbsolutePath();
final static String CHECKSUM_STORE_DIR = new File(TEST_DIR,
"ckm_store." + System.currentTimeMillis()).getAbsolutePath();
final static String CONFIG_FILE = new File(TEST_DIR,
"test-raid.xml").getAbsolutePath();
final static long RELOAD_INTERVAL = 1000;
final static int NUM_DATANODES = 4;
Configuration conf;
String namenode = null;
MiniDFSCluster dfsCluster = null;
String hftp = null;
MiniMRCluster mr = null;
FileSystem fileSys = null;
RaidNode cnode = null;
String jobTrackerName = null;
Random rand = new Random();
static {
ParityFilePair.disableCacheUsedInTestOnly();
}
class TestSendRecoveryTimeInjectionHandler extends InjectionHandler {
@Override
public void _processEvent(InjectionEventI event, Object... args) {
if (event == InjectionEvent.RAID_SEND_RECOVERY_TIME) {
if (cnode == null) {
return;
}
RaidHistogram histogram = (RaidHistogram)args[0];
String p = (String)args[1];
Long value = (Long)args[2];
ArrayList<Point> points =
histogram.getPointsWithGivenRecoveryTime(value);
boolean match = false;
for (Point pt: points) {
if (pt.path.equals(p)) {
match = true;
assertEquals(value, (Long)pt.value);
String trackingUrl = CorruptFileCounterServlet.getTrackingUrl(
(String)args[3], cnode);
assertTrue("Should get a tracking url", trackingUrl.length() > 0);
break;
}
}
assertTrue("We should find path " + p + " in the histogram", match);
}
}
}
public static void verifyMetrics(FileSystem fileSys,
RaidNode cnode, boolean local, long expectedFixedFiles,
long expectedFixBlocks) {
assertTrue("Fewer expected fixed files",
cnode.blockIntegrityMonitor.getNumFilesFixed() >= expectedFixedFiles);
if (!local) {
verifyMetrics(fileSys, cnode, LOGTYPES.OFFLINE_RECONSTRUCTION_BLOCK,
LOGRESULTS.SUCCESS, null, expectedFixBlocks, true);
}
}
public static void verifyMetrics(FileSystem fileSys,
RaidNode cnode, LOGTYPES type, LOGRESULTS result, long expected,
boolean greater) {
verifyMetrics(fileSys, cnode, type, result, null, expected, greater);
}
public static void verifyMetrics(FileSystem fileSys,
RaidNode cnode, LOGTYPES type, LOGRESULTS result, String tag, long expected,
boolean greater) {
String counterName = LogUtils.getCounterName(fileSys, type, result, tag);
Map<String, MetricsTimeVaryingLong> logMetrics = RaidNodeMetrics.getInstance(
RaidNodeMetrics.DEFAULT_NAMESPACE_ID).logMetrics;
String message = "expect " + expected + (greater? " >= ": " = ") + counterName;
long actual = 0L;
synchronized(logMetrics) {
if (expected == 0L) {
if (greater == false) {
assertTrue(message, !logMetrics.containsKey(counterName));
} else {
actual = logMetrics.containsKey(counterName)?logMetrics.get(counterName).getCurrentIntervalValue():
0;
assertTrue(message + " but " + actual, actual >= 0L);
}
} else {
actual = logMetrics.get(counterName).getCurrentIntervalValue();
if (greater == false) {
assertEquals(message + " but " + actual, new Long(expected),
new Long(actual));
} else {
assertTrue(message + " but " + actual, actual >= expected);
}
}
}
}
public static void setChecksumStoreConfig(Configuration conf) {
conf.set(RaidNode.RAID_CHECKSUM_STORE_CLASS_KEY,
"org.apache.hadoop.raid.LocalChecksumStore");
conf.setBoolean(RaidNode.RAID_CHECKSUM_STORE_REQUIRED_KEY, true);
conf.set(LocalChecksumStore.LOCAL_CHECK_STORE_DIR_KEY, CHECKSUM_STORE_DIR);
}
@Test
public void testPriorityEnum() {
// Verify proper ordering of in-place values
assertTrue(Priority.HIGH.higherThan(Priority.LOW));
assertTrue(Priority.LOW.higherThan(Priority.LOWEST));
// Verify ordering of higherThan (guards against bad additions)
int i = 0;
for (Priority p : Priority.values()) {
int j = 0;
for (Priority q : Priority.values()) {
boolean gt = p.higherThan(q);
if (i > j) {
assertTrue("Priority.gt() returned a bad value", gt);
} else {
assertFalse("Priority.gt() returned a bad value", gt);
}
j--;
}
i--;
}
}
@Test
public void testFilterUnfixableFiles() throws Exception {
conf = new Configuration();
dfsCluster = new MiniDFSCluster(conf, NUM_DATANODES, true, null);
dfsCluster.waitActive();
FileSystem fs = dfsCluster.getFileSystem();
Utils.loadTestCodecs(conf);
try {
Configuration testConf = fs.getConf();
BlockIntegrityMonitor blockFixer =
new LocalBlockIntegrityMonitor(testConf, false);
String p1 = "/user/foo/f1";
String p2 = "/user/foo/f2";
String p3 = "/user/foo/bar/f1";
String p4 = "/raid/user/foo";
String p5 = "/raidrs/user/foo/bar";
fs.mkdirs(new Path(p4));
List<String> fileList = new ArrayList<String>();
fileList.add(p1);
fileList.add(p2);
fileList.add(p3);
blockFixer.filterUnreconstructableSourceFiles(fs, fileList.iterator());
// p3 should be filtered out.
assertEquals(2, fileList.size());
Set<String> filtered = new HashSet<String>();
for (String p: fileList) filtered.add(p);
assertFalse("File not filtered", filtered.contains(p3));
fileList.add(p3);
fs.mkdirs(new Path(p5));
blockFixer.filterUnreconstructableSourceFiles(fs, fileList.iterator());
// Nothing is filtered.
assertEquals(3, fileList.size());
} finally {
dfsCluster.shutdown();
}
}
/**
* Test the filtering of trash files from the list of corrupt files.
*/
@Test
public void testTrashFilter() {
List<String> files = new LinkedList<String>();
// Paths that do not match the trash pattern.
String p1 = "/user/raid/raidtest/f1";
String p2 = "/user/.Trash/";
// Paths that match the trash pattern.
String p3 = "/user/raid/.Trash/raidtest/f1";
String p4 = "/user/raid/.Trash/";
String p5 = "/tmp/foo";
files.add(p1);
files.add(p3);
files.add(p4);
files.add(p2);
files.add(p5);
Configuration conf = new Configuration();
RaidUtils.filterTrash(conf, files);
assertEquals("expected 2 non-trash files but got " + files.size(),
2, files.size());
for (String p: files) {
assertTrue("wrong file returned by filterTrash",
p == p1 || p == p2);
}
}
@Test
public void testBlockFixDist() throws Exception {
implBlockFix(false, false);
}
@Test
public void testBlockFixLocal() throws Exception {
implBlockFix(true, false);
}
private void verifyMXBean(RaidNode cnode) throws Exception {
MBeanServer mbs = ManagementFactory.getPlatformMBeanServer();
ObjectName mxbeanName = new ObjectName(
"hadoop:service=RaidNode,name=RaidNodeState");
Long timeSinceLastSuccessfulFix =
(Long) mbs.getAttribute(mxbeanName, "TimeSinceLastSuccessfulFix");
assertNotNull(timeSinceLastSuccessfulFix);
if (timeSinceLastSuccessfulFix == 0) {
assertEquals("No files need to fix", 0,
cnode.blockIntegrityMonitor.approximateNumRecoverableFiles);
}
LOG.info("timeSinceLastSuccessfulFix:" + timeSinceLastSuccessfulFix +
"\t" + "approximateNumRecoverableFiles:" +
cnode.blockIntegrityMonitor.approximateNumRecoverableFiles +
"\t" + "lastSuccessfulFixTime:" +
cnode.blockIntegrityMonitor.lastSuccessfulFixTime);
}
/**
* Create a file with three stripes, corrupt a block each in two stripes,
* and wait for the the file to be fixed.
*/
public void implBlockFix(boolean local, boolean hasChecksumStore) throws Exception {
LOG.info("Test testBlockFix started.");
long blockSize = 8192L;
int stripeLength = 3;
mySetup(stripeLength, -1); // never har
Path file1 = new Path("/user/dhruba/raidtest/file1");
Path destPath = new Path("/destraid/user/dhruba/raidtest");
long crc1 = TestRaidDfs.createTestFilePartialLastBlock(fileSys, file1,
1, 7, blockSize);
long file1Len = fileSys.getFileStatus(file1).getLen();
LOG.info("Test testBlockFix created test files");
// create an instance of the RaidNode
Configuration localConf = new Configuration(conf);
localConf.setInt("raid.blockfix.interval", 1000);
if (local) {
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.LocalBlockIntegrityMonitor");
} else {
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.DistBlockIntegrityMonitor");
InjectionHandler h = new TestSendRecoveryTimeInjectionHandler();
InjectionHandler.set(h);
}
localConf.setLong("raid.blockfix.filespertask", 2L);
if (hasChecksumStore) {
TestBlockFixer.setChecksumStoreConfig(localConf);
}
try {
cnode = RaidNode.createRaidNode(null, localConf);
TestRaidDfs.waitForFileRaided(LOG, fileSys, file1, destPath);
LOG.info("Startup raidnode");
verifyMXBean(cnode);
cnode.stop(); cnode.join();
FileStatus srcStat = fileSys.getFileStatus(file1);
DistributedFileSystem dfs = (DistributedFileSystem)fileSys;
LocatedBlocks locs = RaidDFSUtil.getBlockLocations(
dfs, file1.toUri().getPath(), 0, srcStat.getLen());
String[] corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("no corrupt files expected", 0, corruptFiles.length);
assertEquals("filesFixed() should return 0 before fixing files",
0, cnode.blockIntegrityMonitor.getNumFilesFixed());
// Corrupt blocks in two different stripes. We can fix them.
int[] corruptBlockIdxs = new int[]{0, 4, 6};
for (int idx: corruptBlockIdxs)
corruptBlock(locs.get(idx).getBlock(), dfsCluster);
RaidDFSUtil.reportCorruptBlocks(dfs, file1, corruptBlockIdxs, blockSize);
corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("file not corrupted", 1, corruptFiles.length);
assertEquals("wrong file corrupted",
corruptFiles[0], file1.toUri().getPath());
assertEquals("wrong number of corrupt blocks", 3,
RaidDFSUtil.corruptBlocksInFile(dfs, file1.toUri().getPath(), 0,
srcStat.getLen()).size());
cnode = RaidNode.createRaidNode(null, localConf);
long start = System.currentTimeMillis();
while ((cnode.blockIntegrityMonitor.getNumFilesFixed() < 1 ||
cnode.blockIntegrityMonitor.getNumberOfPoints("/") < 1) &&
System.currentTimeMillis() - start < 120000) {
LOG.info("Test testBlockFix waiting for files to be fixed.");
verifyMXBean(cnode);
Thread.sleep(1000);
}
LOG.info("Files should be fixed");
verifyMXBean(cnode);
assertTrue("Raidnode should record more than 1 point",
cnode.blockIntegrityMonitor.getNumberOfPoints("/") >= 1);
verifyMetrics(fileSys, cnode, local, 1L, corruptBlockIdxs.length);
dfs = getDFS(conf, dfs);
assertTrue("file not fixed",
TestRaidDfs.validateFile(dfs, file1, file1Len, crc1));
LOG.info("Finish checking");
verifyMXBean(cnode);
} catch (Exception e) {
LOG.info("Test testBlockFix Exception " + e +
StringUtils.stringifyException(e));
throw e;
} finally {
myTearDown();
}
LOG.info("Test testBlockFix completed.");
}
/**
* Tests integrity of generated block.
* Create a file and delete a block entirely. Wait for the block to be
* regenerated. Now stop RaidNode and corrupt the generated block.
* Test that corruption in the generated block can be detected by clients.
*/
public void generatedBlockTestCommon(String testName, int blockToCorrupt,
boolean local, boolean hasChecksumStore) throws Exception {
LOG.info("Test " + testName + " started.");
long blockSize = 8192L;
int stripeLength = 3;
mySetup(stripeLength, -1); // never har
Path file1 = new Path("/user/dhruba/raidtest/file1");
Path destPath = new Path("/destraid/user/dhruba/raidtest");
long crc1 = TestRaidDfs.createTestFile(fileSys, file1, 1, 7, blockSize);
long file1Len = fileSys.getFileStatus(file1).getLen();
LOG.info("Test " + testName + " created test files");
// create an instance of the RaidNode
Configuration localConf = new Configuration(conf);
localConf.setInt("raid.blockfix.interval", 1000);
if (local) {
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.LocalBlockIntegrityMonitor");
} else {
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.DistBlockIntegrityMonitor");
}
if (hasChecksumStore) {
TestBlockFixer.setChecksumStoreConfig(localConf);
}
localConf.setLong("raid.blockfix.filespertask", 2L);
try {
cnode = RaidNode.createRaidNode(null, localConf);
TestRaidDfs.waitForFileRaided(LOG, fileSys, file1, destPath);
cnode.stop(); cnode.join();
FileStatus srcStat = fileSys.getFileStatus(file1);
DistributedFileSystem dfs = (DistributedFileSystem)fileSys;
LocatedBlocks locs = RaidDFSUtil.getBlockLocations(
dfs, file1.toUri().getPath(), 0, srcStat.getLen());
String[] corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("no corrupt files expected", 0, corruptFiles.length);
assertEquals("filesFixed() should return 0 before fixing files",
0, cnode.blockIntegrityMonitor.getNumFilesFixed());
corruptBlock(locs.get(0).getBlock(), dfsCluster);
RaidDFSUtil.reportCorruptBlocks(dfs, file1, new int[]{0}, blockSize);
corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("file not corrupted",
1, corruptFiles.length);
assertEquals("wrong file corrupted",
corruptFiles[0], file1.toUri().getPath());
cnode = RaidNode.createRaidNode(null, localConf);
long start = System.currentTimeMillis();
while ((cnode.blockIntegrityMonitor.getNumFilesFixed() < 1 ||
cnode.blockIntegrityMonitor.getNumberOfPoints("/") < 1) &&
System.currentTimeMillis() - start < 120000) {
LOG.info("Test " + testName + " waiting for files to be fixed.");
Thread.sleep(1000);
}
verifyMetrics(fileSys, cnode, local, 1L, 1L);
assertTrue("Raidnode should record more than 1 point",
cnode.blockIntegrityMonitor.getNumberOfPoints("/") >= 1);
// Stop RaidNode
cnode.stop(); cnode.join(); cnode = null;
// The block has successfully been reconstructed.
dfs = getDFS(conf, dfs);
assertTrue("file not fixed",
TestRaidDfs.validateFile(dfs, file1, file1Len, crc1));
// Now corrupt the generated block.
locs = RaidDFSUtil.getBlockLocations(
dfs, file1.toUri().getPath(), 0, srcStat.getLen());
corruptBlock(locs.get(0).getBlock(), dfsCluster);
RaidDFSUtil.reportCorruptBlocks(dfs, file1, new int[]{0}, blockSize);
try {
Thread.sleep(5*1000);
} catch (InterruptedException ignore) {
}
try {
TestRaidDfs.validateFile(dfs, file1, file1Len, crc1);
fail("Expected exception not thrown");
} catch (org.apache.hadoop.fs.ChecksumException ce) {
} catch (org.apache.hadoop.fs.BlockMissingException bme) {
}
} catch (Exception e) {
LOG.info("Test " + testName + " Exception " + e +
StringUtils.stringifyException(e));
throw e;
} finally {
myTearDown();
}
LOG.info("Test " + testName + " completed.");
}
/**
* Tests integrity of generated block.
* Create a file and delete a block entirely. Wait for the block to be
* regenerated. Now stop RaidNode and corrupt the generated block.
* Test that corruption in the generated block can be detected by clients.
*/
@Test
public void testGeneratedBlockDist() throws Exception {
generatedBlockTestCommon("testGeneratedBlock", 3, false, false);
}
/**
* Tests integrity of generated block.
* Create a file and delete a block entirely. Wait for the block to be
* regenerated. Now stop RaidNode and corrupt the generated block.
* Test that corruption in the generated block can be detected by clients.
*/
@Test
public void testGeneratedBlockLocal() throws Exception {
generatedBlockTestCommon("testGeneratedBlock", 3, true, false);
}
/**
* Tests integrity of generated last block.
* Create a file and delete a block entirely. Wait for the block to be
* regenerated. Now stop RaidNode and corrupt the generated block.
* Test that corruption in the generated block can be detected by clients.
*/
@Test
public void testGeneratedLastBlockDist() throws Exception {
generatedBlockTestCommon("testGeneratedLastBlock", 6, false, false);
}
/**
* Tests integrity of generated last block.
* Create a file and delete a block entirely. Wait for the block to be
* regenerated. Now stop RaidNode and corrupt the generated block.
* Test that corruption in the generated block can be detected by clients.
*/
@Test
public void testGeneratedLastBlockLocal() throws Exception {
generatedBlockTestCommon("testGeneratedLastBlock", 6, true, false);
}
@Test
public void testParityBlockFixDist() throws Exception {
implParityBlockFix("testParityBlockFixDist", false, false);
}
@Test
public void testParityBlockFixLocal() throws Exception {
implParityBlockFix("testParityBlockFixLocal", true, false);
}
/**
* Corrupt a parity file and wait for it to get fixed.
*/
public void implParityBlockFix(String testName, boolean local, boolean hasChecksumStore)
throws Exception {
LOG.info("Test " + testName + " started.");
long blockSize = 8192L;
int stripeLength = 3;
mySetup(stripeLength, -1); // never har
Path file1 = new Path("/user/dhruba/raidtest/file1");
Path destPath = new Path("/destraid/user/dhruba/raidtest");
Path parityFile = new Path("/destraid/user/dhruba/raidtest/file1");
TestRaidDfs.createTestFilePartialLastBlock(fileSys, file1,
1, 7, blockSize);
LOG.info("Test " + testName + " created test files");
// create an instance of the RaidNode
Configuration localConf = new Configuration(conf);
localConf.setInt("raid.blockfix.interval", 1000);
if (local) {
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.LocalBlockIntegrityMonitor");
} else {
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.DistBlockIntegrityMonitor");
}
if (hasChecksumStore) {
TestBlockFixer.setChecksumStoreConfig(localConf);
}
localConf.setLong("raid.blockfix.filespertask", 2L);
try {
cnode = RaidNode.createRaidNode(null, localConf);
TestRaidDfs.waitForFileRaided(LOG, fileSys, file1, destPath);
cnode.stop(); cnode.join();
long parityCRC = RaidDFSUtil.getCRC(fileSys, parityFile);
FileStatus parityStat = fileSys.getFileStatus(parityFile);
DistributedFileSystem dfs = (DistributedFileSystem)fileSys;
LocatedBlocks locs = RaidDFSUtil.getBlockLocations(
dfs, parityFile.toUri().getPath(), 0, parityStat.getLen());
String[] corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("no corrupt files expected", 0, corruptFiles.length);
assertEquals("filesFixed() should return 0 before fixing files",
0, cnode.blockIntegrityMonitor.getNumFilesFixed());
// Corrupt parity blocks for different stripes.
int[] corruptBlockIdxs = new int[]{0, 1, 2};
for (int idx: corruptBlockIdxs)
corruptBlock(locs.get(idx).getBlock(), dfsCluster);
RaidDFSUtil.reportCorruptBlocks(dfs, parityFile, corruptBlockIdxs, blockSize);
corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("file not corrupted",
1, corruptFiles.length);
assertEquals("wrong file corrupted",
corruptFiles[0], parityFile.toUri().getPath());
if (!local) {
assertFalse(dfs.exists(new Path("blockfixer")));
}
cnode = RaidNode.createRaidNode(null, localConf);
long start = System.currentTimeMillis();
while ((cnode.blockIntegrityMonitor.getNumFilesFixed() < 1 ||
cnode.blockIntegrityMonitor.getNumberOfPoints("/") < 1) &&
System.currentTimeMillis() - start < 120000) {
LOG.info("Test " + testName + " waiting for files to be fixed.");
Thread.sleep(1000);
}
assertTrue("Raidnode should record more than 1 point",
cnode.blockIntegrityMonitor.getNumberOfPoints("/") >= 1);
long checkCRC = RaidDFSUtil.getCRC(fileSys, parityFile);
assertEquals("file not fixed",
parityCRC, checkCRC);
cnode.stop(); cnode.join();
if (!local) {
assertTrue("blockfixer will create /user/username/blockfixer",
dfs.exists(new Path("blockfixer")));
}
cnode = RaidNode.createRaidNode(null, localConf);
if (!local) {
assertFalse("Restarting raidnode will cleanup job dir",
dfs.exists(new Path("blockfixer")));
}
} catch (Exception e) {
LOG.info("Test " + testName + " Exception " + e +
StringUtils.stringifyException(e));
throw e;
} finally {
myTearDown();
}
LOG.info("Test " + testName + " completed.");
}
@Test
public void testParityHarBlockFixDist() throws Exception {
implParityHarBlockFix("testParityHarBlockFixDist", false);
}
@Test
public void testParityHarBlockFixLocal() throws Exception {
implParityHarBlockFix("testParityHarBlockFixLocal", true);
}
private void implParityHarBlockFix(String testName, boolean local)
throws Exception {
LOG.info("Test " + testName + " started.");
long blockSize = 8192L;
int stripeLength = 3;
mySetup(stripeLength, 0); // Time before har = 0 days.
Path file1 = new Path("/user/dhruba/raidtest/file1");
// Parity file will have 7 blocks.
TestRaidDfs.createTestFilePartialLastBlock(fileSys, file1,
1, 20, blockSize);
LOG.info("Test " + testName + " created test files");
// create an instance of the RaidNode
Configuration localConf = new Configuration(conf);
localConf.setInt("raid.blockfix.interval", 1000);
localConf.setInt(RaidNode.RAID_PARITY_HAR_THRESHOLD_DAYS_KEY, 0);
if (local) {
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.LocalBlockIntegrityMonitor");
} else {
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.DistBlockIntegrityMonitor");
}
localConf.setLong("raid.blockfix.filespertask", 2L);
try {
cnode = RaidNode.createRaidNode(null, localConf);
Path harDirectory =
new Path("/destraid/user/dhruba/raidtest/raidtest" +
RaidNode.HAR_SUFFIX);
long start = System.currentTimeMillis();
while (System.currentTimeMillis() - start < 1000 * 120) {
if (fileSys.exists(harDirectory)) {
break;
}
LOG.info("Test " + testName + " waiting for har");
Thread.sleep(1000);
}
assertEquals(true, fileSys.exists(harDirectory));
Path partFile = new Path(harDirectory, "part-0");
long partCRC = RaidDFSUtil.getCRC(fileSys, partFile);
FileStatus partStat = fileSys.getFileStatus(partFile);
DistributedFileSystem dfs = (DistributedFileSystem)fileSys;
LocatedBlocks locs = RaidDFSUtil.getBlockLocations(
dfs, partFile.toUri().getPath(), 0, partStat.getLen());
assertEquals("wrong number of har blocks",
7, locs.getLocatedBlocks().size());
cnode.stop(); cnode.join();
String[] corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("no corrupt files expected", 0, corruptFiles.length);
assertEquals("filesFixed() should return 0 before fixing files",
0, cnode.blockIntegrityMonitor.getNumFilesFixed());
// Corrupt parity blocks for different stripes.
int[] corruptBlockIdxs = new int[]{0, 1, 6};
for (int idx: corruptBlockIdxs)
corruptBlock(locs.get(idx).getBlock(), dfsCluster);
RaidDFSUtil.reportCorruptBlocks(dfs, partFile, corruptBlockIdxs,
partStat.getBlockSize());
corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("file not corrupted", 1, corruptFiles.length);
assertEquals("wrong file corrupted",
corruptFiles[0], partFile.toUri().getPath());
cnode = RaidNode.createRaidNode(null, localConf);
start = System.currentTimeMillis();
while ((cnode.blockIntegrityMonitor.getNumFilesFixed() < 1 ||
cnode.blockIntegrityMonitor.getNumberOfPoints("/") < 1) &&
System.currentTimeMillis() - start < 120000) {
LOG.info("Test " + testName + " waiting for files to be fixed.");
Thread.sleep(1000);
}
assertTrue("Raidnode should record more than 1 point",
cnode.blockIntegrityMonitor.getNumberOfPoints("/") >= 1);
long checkCRC = RaidDFSUtil.getCRC(fileSys, partFile);
assertEquals("file not fixed",
partCRC, checkCRC);
} catch (Exception e) {
LOG.info("Test " + testName + " Exception " + e +
StringUtils.stringifyException(e));
throw e;
} finally {
myTearDown();
}
LOG.info("Test " + testName + " completed.");
}
/**
* tests that we can have 2 concurrent jobs fixing files
* (dist block fixer)
*/
@Test
public void testConcurrentJobs() throws Exception {
LOG.info("Test testConcurrentJobs started.");
long blockSize = 8192L;
int stripeLength = 3;
mySetup(stripeLength, -1); // never har
Path file1 = new Path("/user/dhruba/raidtest/concurrentjobfile1");
Path file2 = new Path("/user/dhruba/raidtest/concurrentjobfile2");
Path destPath = new Path("/destraid/user/dhruba/raidtest");
long crc1 = TestRaidDfs.createTestFilePartialLastBlock(fileSys, file1,
1, 20, blockSize);
long crc2 = TestRaidDfs.createTestFilePartialLastBlock(fileSys, file2,
1, 20, blockSize);
long file1Len = fileSys.getFileStatus(file1).getLen();
long file2Len = fileSys.getFileStatus(file2).getLen();
LOG.info("Test testConcurrentJobs created test files");
// create an instance of the RaidNode
Configuration localConf = new Configuration(conf);
localConf.setInt("raid.blockfix.interval", 1000);
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.DistBlockIntegrityMonitor");
localConf.setLong("raid.blockfix.filespertask", 2L);
TestBlockFixer.setChecksumStoreConfig(localConf);
try {
cnode = RaidNode.createRaidNode(null, localConf);
TestRaidDfs.waitForFileRaided(LOG, fileSys, file1, destPath);
TestRaidDfs.waitForFileRaided(LOG, fileSys, file2, destPath);
cnode.stop(); cnode.join();
FileStatus file1Stat = fileSys.getFileStatus(file1);
FileStatus file2Stat = fileSys.getFileStatus(file2);
DistributedFileSystem dfs = (DistributedFileSystem)fileSys;
LocatedBlocks file1Loc =
RaidDFSUtil.getBlockLocations(dfs, file1.toUri().getPath(),
0, file1Stat.getLen());
LocatedBlocks file2Loc =
RaidDFSUtil.getBlockLocations(dfs, file2.toUri().getPath(),
0, file2Stat.getLen());
String[] corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("no corrupt files expected", 0, corruptFiles.length);
assertEquals("filesFixed() should return 0 before fixing files",
0, cnode.blockIntegrityMonitor.getNumFilesFixed());
// corrupt file1
int[] corruptBlockIdxs = new int[]{0, 4, 6};
for (int idx: corruptBlockIdxs) {
corruptBlock(file1Loc.get(idx).getBlock(), dfsCluster);
RaidDFSUtil.reportCorruptBlocksToNN(dfs,
new LocatedBlock[] {file1Loc.get(idx)});
}
RaidDFSUtil.reportCorruptBlocks(dfs, file1, corruptBlockIdxs, blockSize);
cnode = RaidNode.createRaidNode(null, localConf);
DistBlockIntegrityMonitor blockFixer = (DistBlockIntegrityMonitor) cnode.blockIntegrityMonitor;
long start = System.currentTimeMillis();
while (blockFixer.jobsRunning() < 1 &&
System.currentTimeMillis() - start < 240000) {
LOG.info("Test testBlockFix waiting for fixing job 1 to start");
Thread.sleep(10);
}
assertEquals("job 1 not running", 1, blockFixer.jobsRunning());
// corrupt file2
for (int idx: corruptBlockIdxs) {
corruptBlock(file2Loc.get(idx).getBlock(), dfsCluster);
RaidDFSUtil.reportCorruptBlocksToNN(dfs,
new LocatedBlock[] {file2Loc.get(idx)});
}
corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("2 corrupt files expected", 2, corruptFiles.length);
while (blockFixer.jobsRunning() < 2 &&
System.currentTimeMillis() - start < 240000) {
LOG.info("Test testBlockFix waiting for fixing job 2 to start");
Thread.sleep(10);
}
assertTrue(blockFixer.jobsRunning() >= 2);
while ((blockFixer.getNumFilesFixed() < 2 ||
blockFixer.getNumberOfPoints("/") < 2) &&
System.currentTimeMillis() - start < 240000) {
LOG.info("Test testBlockFix waiting for files to be fixed.");
Thread.sleep(10);
}
// wait for all the jobs to finish
while (blockFixer.jobsRunning() > 0 &&
System.currentTimeMillis() - start < 240000) {
LOG.info("Test testBlockFix waiting for block fixer jobs to finish.");
Thread.sleep(10);
}
dfs = getDFS(conf, dfs);
try {
Thread.sleep(5*1000);
} catch (InterruptedException ignore) {
}
assertTrue("files not fixed", blockFixer.getNumFilesFixed() >= 2);
assertTrue("fixed files not recorded",
blockFixer.getNumberOfPoints("/") >= 2);
verifyMetrics(fileSys, cnode, LOGTYPES.OFFLINE_RECONSTRUCTION_BLOCK,
LOGRESULTS.SUCCESS, corruptBlockIdxs.length * 2, true);
assertTrue("file not fixed",
TestRaidDfs.validateFile(dfs, file1, file1Len, crc1));
assertTrue("file not fixed",
TestRaidDfs.validateFile(dfs, file2, file2Len, crc2));
} catch (Exception e) {
LOG.info("Test testConcurrentJobs exception " + e +
StringUtils.stringifyException(e));
throw e;
} finally {
myTearDown();
}
}
/**
* tests that the distributed block fixer obeys
* the limit on how many jobs to submit simultaneously.
*/
@Test
public void testMaxPendingJobs() throws Exception {
LOG.info("Test testMaxPendingJobs started.");
long blockSize = 8192L;
int stripeLength = 3;
mySetup(stripeLength, -1); // never har
Path file1 = new Path("/user/dhruba/raidtest/file1");
Path file2 = new Path("/user/dhruba/raidtest/file2");
Path destPath = new Path("/destraid/user/dhruba/raidtest");
long crc1 = TestRaidDfs.createTestFilePartialLastBlock(fileSys, file1,
1, 20, blockSize);
long crc2 = TestRaidDfs.createTestFilePartialLastBlock(fileSys, file2,
1, 20, blockSize);
long file1Len = fileSys.getFileStatus(file1).getLen();
long file2Len = fileSys.getFileStatus(file2).getLen();
LOG.info("Test testMaxPendingJobs created test files");
// create an instance of the RaidNode
Configuration localConf = new Configuration(conf);
localConf.setInt("raid.blockfix.interval", 1000);
localConf.set("raid.blockfix.classname",
"org.apache.hadoop.raid.DistBlockIntegrityMonitor");
localConf.setLong("raid.blockfix.filespertask", 2L);
localConf.setLong("raid.blockfix.maxpendingjobs", 1L);
TestBlockFixer.setChecksumStoreConfig(localConf);
try {
cnode = RaidNode.createRaidNode(null, localConf);
TestRaidDfs.waitForFileRaided(LOG, fileSys, file1, destPath);
TestRaidDfs.waitForFileRaided(LOG, fileSys, file2, destPath);
cnode.stop(); cnode.join();
FileStatus file1Stat = fileSys.getFileStatus(file1);
FileStatus file2Stat = fileSys.getFileStatus(file2);
DistributedFileSystem dfs = (DistributedFileSystem)fileSys;
LocatedBlocks file1Loc =
RaidDFSUtil.getBlockLocations(dfs, file1.toUri().getPath(),
0, file1Stat.getLen());
LocatedBlocks file2Loc =
RaidDFSUtil.getBlockLocations(dfs, file2.toUri().getPath(),
0, file2Stat.getLen());
String[] corruptFiles = DFSUtil.getCorruptFiles(dfs);
assertEquals("no corrupt files expected", 0, corruptFiles.length);
assertEquals("filesFixed() should return 0 before fixing files",
0, cnode.blockIntegrityMonitor.getNumFilesFixed());
// corrupt file1
int[] corruptBlockIdxs = new int[]{0, 4, 6};
for (int idx: corruptBlockIdxs)
corruptBlock(file1Loc.get(idx).getBlock(), dfsCluster);
RaidDFSUtil.reportCorruptBlocks(dfs, file1, corruptBlockIdxs, blockSize);
corruptFiles = DFSUtil.getCorruptFiles(dfs);
cnode = RaidNode.createRaidNode(null, localConf);
DistBlockIntegrityMonitor blockFixer = (DistBlockIntegrityMonitor) cnode.blockIntegrityMonitor;
long start = System.currentTimeMillis();
while (blockFixer.jobsRunning() < 1 &&
System.currentTimeMillis() - start < 240000) {
LOG.info("Test testBlockFix waiting for fixing job 1 to start");
Thread.sleep(10);
}
assertEquals("job not running", 1, blockFixer.jobsRunning());
// corrupt file2
for (int idx: corruptBlockIdxs)
corruptBlock(file2Loc.get(idx).getBlock(), dfsCluster);
RaidDFSUtil.reportCorruptBlocks(dfs, file2, corruptBlockIdxs, blockSize);
corruptFiles = DFSUtil.getCorruptFiles(dfs);
// wait until both files are fixed
while ((blockFixer.getNumFilesFixed() < 2 ||
blockFixer.getNumberOfPoints("/") < 2) &&
System.currentTimeMillis() - start < 240000) {
// make sure the block fixer does not start a second job while
// the first one is still running
assertTrue("too many jobs running", blockFixer.jobsRunning() <= 1);
Thread.sleep(10);
}
assertTrue("files not fixed", blockFixer.getNumFilesFixed() >= 2);
assertTrue("files fixed not record",
blockFixer.getNumberOfPoints("/") >= 2);
verifyMetrics(fileSys, cnode, LOGTYPES.OFFLINE_RECONSTRUCTION_BLOCK,
LOGRESULTS.SUCCESS, corruptBlockIdxs.length * 2, true);
dfs = getDFS(conf, dfs);
try {
Thread.sleep(5*1000);
} catch (InterruptedException ignore) {
}
assertTrue("file not fixed",
TestRaidDfs.validateFile(dfs, file1, file1Len, crc1));
assertTrue("file not fixed",
TestRaidDfs.validateFile(dfs, file2, file2Len, crc2));
} catch (Exception e) {
LOG.info("Test testMaxPendingJobs exception " + e +
StringUtils.stringifyException(e));
throw e;
} finally {
myTearDown();
}
}
static class FakeDistBlockIntegrityMonitor extends DistBlockIntegrityMonitor {
Map<String, List<String>> submittedJobs =
new HashMap<String, List<String>>();
FakeDistBlockIntegrityMonitor(Configuration conf) throws Exception {
super(conf);
}
@Override
void submitJob(Job job, List<String> filesInJob, Priority priority,
Map<Job, List<LostFileInfo>> jobIndex,
Map<JobID, TrackingUrlInfo> idToTrackingUrlMap) {
LOG.info("Job " + job.getJobName() + " was submitted ");
submittedJobs.put(job.getJobName(), filesInJob);
}
}
public void testMultiplePriorities() throws Exception {
LOG.info("Test testMultiplePriorities started.");
Path srcFile = new Path("/home/test/file1");
int repl = 1;
int numBlocks = 8;
long blockSize = 16384;
int stripeLength = 3;
Path destPath = new Path("/destraidrs");
mySetup(stripeLength, -1); // never har
Codec codec = Codec.getCodec("rs");
LOG.info("Starting testMultiplePriorities");
try {
// Create test file and raid it.
TestRaidDfs.createTestFilePartialLastBlock(
fileSys, srcFile, repl, numBlocks, blockSize);
FileStatus stat = fileSys.getFileStatus(srcFile);
RaidNode.doRaid(conf, stat,
destPath, codec, new RaidNode.Statistics(), RaidUtils.NULL_PROGRESSABLE,
false, repl, repl);
// Corrupt first block of file.
int blockIdxToCorrupt = 1;
LOG.info("Corrupt block " + blockIdxToCorrupt + " of file " + srcFile);
LocatedBlocks locations = getBlockLocations(srcFile, stat.getLen());
corruptBlock(locations.get(blockIdxToCorrupt).getBlock(),
dfsCluster);
RaidDFSUtil.reportCorruptBlocks(fileSys, srcFile, new int[]{1}, blockSize);
// Create Block Fixer and fix.
FakeDistBlockIntegrityMonitor distBlockFixer = new FakeDistBlockIntegrityMonitor(conf);
assertEquals(0, distBlockFixer.submittedJobs.size());
// waiting for one job to submit
long startTime = System.currentTimeMillis();
while (System.currentTimeMillis() - startTime < 120000 &&
distBlockFixer.submittedJobs.size() == 0) {
distBlockFixer.getCorruptionMonitor().checkAndReconstructBlocks();
LOG.info("Waiting for jobs to submit");
Thread.sleep(10000);
}
int submittedJob = distBlockFixer.submittedJobs.size();
LOG.info("Already Submitted " + submittedJob + " jobs");
assertTrue("Should submit more than 1 jobs", submittedJob >= 1);
// Corrupt one more block.
blockIdxToCorrupt = 4;
LOG.info("Corrupt block " + blockIdxToCorrupt + " of file " + srcFile);
locations = getBlockLocations(srcFile, stat.getLen());
corruptBlock(locations.get(blockIdxToCorrupt).getBlock(),
dfsCluster);
RaidDFSUtil.reportCorruptBlocks(fileSys, srcFile, new int[]{4}, blockSize);
// A new job should be submitted since two blocks are corrupt.
startTime = System.currentTimeMillis();
while (System.currentTimeMillis() - startTime < 120000 &&
distBlockFixer.submittedJobs.size() == submittedJob) {
distBlockFixer.getCorruptionMonitor().checkAndReconstructBlocks();
LOG.info("Waiting for more jobs to submit");
Thread.sleep(10000);
}
LOG.info("Already Submitted " + distBlockFixer.submittedJobs.size() + " jobs");
assertTrue("Should submit more than 1 jobs",
distBlockFixer.submittedJobs.size() - submittedJob >= 1);
} finally {
myTearDown();
}
}
public static DistributedFileSystem getDFS(
Configuration conf, FileSystem dfs) throws IOException {
Configuration clientConf = new Configuration(conf);
clientConf.set("fs.hdfs.impl",
"org.apache.hadoop.hdfs.DistributedFileSystem");
clientConf.setBoolean("fs.hdfs.impl.disable.cache", true);
URI dfsUri = dfs.getUri();
FileSystem.closeAll();
return (DistributedFileSystem) FileSystem.get(dfsUri, clientConf);
}
private void mySetup(int stripeLength, int timeBeforeHar) throws Exception {
if (System.getProperty("hadoop.log.dir") == null) {
String base = new File(".").getAbsolutePath();
System.setProperty("hadoop.log.dir", new Path(base).toString() + "/logs");
}
new File(TEST_DIR).mkdirs(); // Make sure data directory exists
conf = new Configuration();
conf.set("raid.config.file", CONFIG_FILE);
conf.setBoolean("raid.config.reload", true);
conf.setLong("raid.config.reload.interval", RELOAD_INTERVAL);
// scan all policies once every 5 second
conf.setLong("raid.policy.rescan.interval", 5000);
// do not use map-reduce cluster for Raiding
conf.set("raid.classname", "org.apache.hadoop.raid.LocalRaidNode");
conf.set("raid.server.address", "localhost:" + MiniDFSCluster.getFreePort());
conf.set("mapred.raid.http.address", "localhost:0");
// Make sure initial repl is smaller than NUM_DATANODES
conf.setInt(RaidNode.RAID_PARITY_INITIAL_REPL_KEY, 1);
Utils.loadTestCodecs(conf, stripeLength, 1, 3, "/destraid", "/destraidrs");
conf.setBoolean("dfs.permissions", false);
dfsCluster = new MiniDFSCluster(conf, NUM_DATANODES, true, null);
dfsCluster.waitActive();
fileSys = dfsCluster.getFileSystem();
namenode = fileSys.getUri().toString();
FileSystem.setDefaultUri(conf, namenode);
mr = new MiniMRCluster(4, namenode, 3);
jobTrackerName = "localhost:" + mr.getJobTrackerPort();
hftp = "hftp://localhost.localdomain:" + dfsCluster.getNameNodePort();
FileSystem.setDefaultUri(conf, namenode);
conf.set("mapred.job.tracker" + "." + DistBlockIntegrityMonitor.BLOCKFIXER, jobTrackerName);
FileWriter fileWriter = new FileWriter(CONFIG_FILE);
fileWriter.write("<?xml version=\"1.0\"?>\n");
String str = "<configuration> " +
"<policy name = \"RaidTest1\"> " +
"<srcPath prefix=\"/user/dhruba/raidtest\"/> " +
"<codecId>xor</codecId> " +
"<destPath> /destraid</destPath> " +
"<property> " +
"<name>targetReplication</name> " +
"<value>1</value> " +
"<description>after RAIDing, decrease the replication factor of a file to this value." +
"</description> " +
"</property> " +
"<property> " +
"<name>metaReplication</name> " +
"<value>1</value> " +
"<description> replication factor of parity file" +
"</description> " +
"</property> " +
"<property> " +
"<name>modTimePeriod</name> " +
"<value>2000</value> " +
"<description> time (milliseconds) after a file is modified to make it " +
"a candidate for RAIDing " +
"</description> " +
"</property> ";
if (timeBeforeHar >= 0) {
str +=
"<property> " +
"<name>time_before_har</name> " +
"<value>" + timeBeforeHar + "</value> " +
"<description> amount of time waited before har'ing parity files" +
"</description> " +
"</property> ";
}
str +=
"</policy>" +
"</configuration>";
fileWriter.write(str);
fileWriter.close();
}
private void myTearDown() throws Exception {
if (cnode != null) { cnode.stop(); cnode.join(); }
if (mr != null) { mr.shutdown(); }
if (dfsCluster != null) { dfsCluster.shutdown(); }
InjectionHandler.clear();
}
private LocatedBlocks getBlockLocations(Path file, long length)
throws IOException {
DistributedFileSystem dfs = (DistributedFileSystem) fileSys;
return dfs.getClient().namenode.getBlockLocations(file.toString(),
0, length);
}
static void corruptBlock(Block block, MiniDFSCluster dfs) throws IOException {
boolean corrupted = false;
for (int i = 0; i < NUM_DATANODES; i++) {
corrupted |= TestDatanodeBlockScanner.corruptReplica(block, i, dfs);
}
assertTrue("could not corrupt block", corrupted);
}
}