Apache
/
hadoop
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

HADOOP-18023. Allow cp command to run with multi threads. (#3721) 

(cherry picked from commit 932a78fe38)
(cherry picked from commit 2f3e186978)
This commit is contained in:
smarthan 2021-11-29 20:45:08 +08:00 committed by S O'Donnell
parent d15afa3da8
commit 4e9b9adb96
4 changed files with 275 additions and 32 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

39
hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/shell/CopyCommands.java
View File

@ -146,33 +146,40 @@ class CopyCommands {
} }
} }
static class Cp extends CommandWithDestination { static class Cp extends CopyCommandWithMultiThread {
public static final String NAME = "cp"; public static final String NAME = "cp";
public static final String USAGE = public static final String USAGE =
"[-f] [-p | -p[topax]] [-d] <src> ... <dst>"; "[-f] [-p | -p[topax]] [-d] [-t <thread count>]"
+ " [-q <thread pool queue size>] <src> ... <dst>";
public static final String DESCRIPTION = public static final String DESCRIPTION =
"Copy files that match the file pattern <src> to a " + "Copy files that match the file pattern <src> to a destination."
"destination. When copying multiple files, the destination " + + " When copying multiple files, the destination must be a "
"must be a directory. Passing -p preserves status " + + "directory.\nFlags :\n"
"[topax] (timestamps, ownership, permission, ACLs, XAttr). " + + " -p[topax] : Preserve file attributes [topx] (timestamps, "
"If -p is specified with no <arg>, then preserves " + + "ownership, permission, ACL, XAttr). If -p is specified with "
"timestamps, ownership, permission. If -pa is specified, " + + "no arg, then preserves timestamps, ownership, permission. "
"then preserves permission also because ACL is a super-set of " + + "If -pa is specified, then preserves permission also because "
"permission. Passing -f overwrites the destination if it " + + "ACL is a super-set of permission. Determination of whether raw "
"already exists. raw namespace extended attributes are preserved " + + "namespace extended attributes are preserved is independent of "
"if (1) they are supported (HDFS only) and, (2) all of the source and " + + "the -p flag.\n"
"target pathnames are in the /.reserved/raw hierarchy. raw namespace " + + " -f : Overwrite the destination if it already exists.\n"
"xattr preservation is determined solely by the presence (or absence) " + + " -d : Skip creation of temporary file(<dst>._COPYING_).\n"
"of the /.reserved/raw prefix and not by the -p option. Passing -d "+ + " -t <thread count> : Number of threads to be used, "
"will skip creation of temporary file(<dst>._COPYING_).\n"; + "default is 1.\n"
+ " -q <thread pool queue size> : Thread pool queue size to be "
+ "used, default is 1024.\n";
@Override @Override
protected void processOptions(LinkedList<String> args) throws IOException { protected void processOptions(LinkedList<String> args) throws IOException {
popPreserveOption(args); popPreserveOption(args);
CommandFormat cf = new CommandFormat(2, Integer.MAX_VALUE, "f", "d"); CommandFormat cf = new CommandFormat(2, Integer.MAX_VALUE, "f", "d");
cf.addOptionWithValue("t");
cf.addOptionWithValue("q");
cf.parse(args); cf.parse(args);
setDirectWrite(cf.getOpt("d")); setDirectWrite(cf.getOpt("d"));
setOverwrite(cf.getOpt("f")); setOverwrite(cf.getOpt("f"));
setThreadCount(cf.getOptValue("t"));
setThreadPoolQueueSize(cf.getOptValue("q"));
// should have a -r option // should have a -r option
setRecursive(true); setRecursive(true);
getRemoteDestination(args); getRemoteDestination(args);

11
hadoop-common-project/hadoop-common/src/site/markdown/FileSystemShell.md
View File

@ -170,7 +170,7 @@ Returns 0 on success and -1 on error.
cp cp
---- ----
Usage: `hadoop fs -cp [-f] [-p | -p[topax]] URI [URI ...] <dest> ` Usage: `hadoop fs -cp [-f] [-p | -p[topax]] [-t <thread count>] [-q <thread pool queue size>] URI [URI ...] <dest>`
Copy files from source to destination. This command allows multiple sources as well in which case the destination must be a directory. Copy files from source to destination. This command allows multiple sources as well in which case the destination must be a directory.
@ -178,13 +178,18 @@ Copy files from source to destination. This command allows multiple sources as w
Options: Options:
* The -f option will overwrite the destination if it already exists. * `-f` : Overwrite the destination if it already exists.
* The -p option will preserve file attributes [topx] (timestamps, ownership, permission, ACL, XAttr). If -p is specified with no *arg*, then preserves timestamps, ownership, permission. If -pa is specified, then preserves permission also because ACL is a super-set of permission. Determination of whether raw namespace extended attributes are preserved is independent of the -p flag. * `-d` : Skip creation of temporary file with the suffix `._COPYING_`.
* `-p` : Preserve file attributes [topx] (timestamps, ownership, permission, ACL, XAttr). If -p is specified with no *arg*, then preserves timestamps, ownership, permission. If -pa is specified, then preserves permission also because ACL is a super-set of permission. Determination of whether raw namespace extended attributes are preserved is independent of the -p flag.
* `-t <thread count>` : Number of threads to be used, default is 1. Useful when copying directories containing more than 1 file.
* `-q <thread pool queue size>` : Thread pool queue size to be used, default is 1024. It takes effect only when thread count greater than 1.
Example: Example:
* `hadoop fs -cp /user/hadoop/file1 /user/hadoop/file2` * `hadoop fs -cp /user/hadoop/file1 /user/hadoop/file2`
* `hadoop fs -cp /user/hadoop/file1 /user/hadoop/file2 /user/hadoop/dir` * `hadoop fs -cp /user/hadoop/file1 /user/hadoop/file2 /user/hadoop/dir`
* `hadoop fs -cp -t 5 /user/hadoop/file1 /user/hadoop/file2 /user/hadoop/dir`
* `hadoop fs -cp -t 10 -q 2048 /user/hadoop/file1 /user/hadoop/file2 /user/hadoop/dir`
Exit Code: Exit Code:

210
hadoop-common-project/hadoop-common/src/test/java/org/apache/hadoop/fs/shell/TestCpCommand.java Normal file
View File

@ -0,0 +1,210 @@
/**
* 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.fs.shell;
import java.io.IOException;
import java.util.LinkedList;
import java.util.concurrent.ThreadPoolExecutor;
import org.junit.AfterClass;
import org.junit.Assert;
import org.junit.Before;
import org.junit.BeforeClass;
import org.junit.Test;
import org.apache.commons.lang3.RandomStringUtils;
import org.apache.commons.lang3.RandomUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FSDataOutputStream;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.FileSystemTestHelper;
import org.apache.hadoop.fs.LocalFileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.fs.shell.CopyCommands.Cp;
import static org.apache.hadoop.fs.shell.CopyCommandWithMultiThread.DEFAULT_QUEUE_SIZE;
import static org.junit.Assert.assertEquals;
public class TestCpCommand {
private static final String FROM_DIR_NAME = "fromDir";
private static final String TO_DIR_NAME = "toDir";
private static FileSystem fs;
private static Path testDir;
private static Configuration conf;
private Path dir = null;
private int numFiles = 0;
private static int initialize(Path dir) throws Exception {
fs.mkdirs(dir);
Path fromDirPath = new Path(dir, FROM_DIR_NAME);
fs.mkdirs(fromDirPath);
Path toDirPath = new Path(dir, TO_DIR_NAME);
fs.mkdirs(toDirPath);
int numTotalFiles = 0;
int numDirs = RandomUtils.nextInt(0, 5);
for (int dirCount = 0; dirCount < numDirs; ++dirCount) {
Path subDirPath = new Path(fromDirPath, "subdir" + dirCount);
fs.mkdirs(subDirPath);
int numFiles = RandomUtils.nextInt(0, 10);
for (int fileCount = 0; fileCount < numFiles; ++fileCount) {
numTotalFiles++;
Path subFile = new Path(subDirPath, "file" + fileCount);
fs.createNewFile(subFile);
FSDataOutputStream output = fs.create(subFile, true);
for (int i = 0; i < 100; ++i) {
output.writeInt(i);
output.writeChar('\n');
}
output.close();
}
}
return numTotalFiles;
}
@BeforeClass
public static void init() throws Exception {
conf = new Configuration(false);
conf.set("fs.file.impl", LocalFileSystem.class.getName());
fs = FileSystem.getLocal(conf);
testDir = new FileSystemTestHelper().getTestRootPath(fs);
// don't want scheme on the path, just an absolute path
testDir = new Path(fs.makeQualified(testDir).toUri().getPath());
FileSystem.setDefaultUri(conf, fs.getUri());
fs.setWorkingDirectory(testDir);
}
@AfterClass
public static void cleanup() throws Exception {
fs.delete(testDir, true);
fs.close();
}
private void run(CopyCommandWithMultiThread cmd, String... args) {
cmd.setConf(conf);
assertEquals(0, cmd.run(args));
}
@Before
public void initDirectory() throws Exception {
dir = new Path("dir" + RandomStringUtils.randomNumeric(4));
numFiles = initialize(dir);
}
@Test(timeout = 10000)
public void testCp() throws Exception {
MultiThreadedCp copy = new MultiThreadedCp(1, DEFAULT_QUEUE_SIZE, 0);
run(copy, new Path(dir, FROM_DIR_NAME).toString(),
new Path(dir, TO_DIR_NAME).toString());
assert copy.getExecutor() == null;
}
@Test(timeout = 10000)
public void testCpWithThreads() {
run(new MultiThreadedCp(5, DEFAULT_QUEUE_SIZE, numFiles), "-t", "5",
new Path(dir, FROM_DIR_NAME).toString(),
new Path(dir, TO_DIR_NAME).toString());
}
@Test(timeout = 10000)
public void testCpWithThreadWrong() {
run(new MultiThreadedCp(1, DEFAULT_QUEUE_SIZE, 0), "-t", "0",
new Path(dir, FROM_DIR_NAME).toString(),
new Path(dir, TO_DIR_NAME).toString());
}
@Test(timeout = 10000)
public void testCpWithThreadsAndQueueSize() {
int queueSize = 256;
run(new MultiThreadedCp(5, queueSize, numFiles), "-t", "5", "-q",
Integer.toString(queueSize),
new Path(dir, FROM_DIR_NAME).toString(),
new Path(dir, TO_DIR_NAME).toString());
}
@Test(timeout = 10000)
public void testCpWithThreadsAndQueueSizeWrong() {
int queueSize = 0;
run(new MultiThreadedCp(5, DEFAULT_QUEUE_SIZE, numFiles), "-t", "5", "-q",
Integer.toString(queueSize),
new Path(dir, FROM_DIR_NAME).toString(),
new Path(dir, TO_DIR_NAME).toString());
}
@Test(timeout = 10000)
public void testCpSingleFile() throws Exception {
Path fromDirPath = new Path(dir, FROM_DIR_NAME);
Path subFile = new Path(fromDirPath, "file0");
fs.createNewFile(subFile);
FSDataOutputStream output = fs.create(subFile, true);
for (int i = 0; i < 100; ++i) {
output.writeInt(i);
output.writeChar('\n');
}
output.close();
MultiThreadedCp copy = new MultiThreadedCp(5, DEFAULT_QUEUE_SIZE, 0);
run(copy, "-t", "5", subFile.toString(),
new Path(dir, TO_DIR_NAME).toString());
assert copy.getExecutor() == null;
}
private static class MultiThreadedCp extends Cp {
public static final String NAME = "multiThreadCp";
private final int expectedThreads;
private final int expectedQueuePoolSize;
private final int expectedCompletedTaskCount;
MultiThreadedCp(int expectedThreads, int expectedQueuePoolSize,
int expectedCompletedTaskCount) {
this.expectedThreads = expectedThreads;
this.expectedQueuePoolSize = expectedQueuePoolSize;
this.expectedCompletedTaskCount = expectedCompletedTaskCount;
}
@Override
protected void processArguments(LinkedList<PathData> args)
throws IOException {
// Check if the number of threads are same as expected
Assert.assertEquals(expectedThreads, getThreadCount());
// Check if the queue pool size of executor is same as expected
Assert.assertEquals(expectedQueuePoolSize, getThreadPoolQueueSize());
super.processArguments(args);
if (isMultiThreadNecessary(args)) {
// Once the copy is complete, check following
// 1) number of completed tasks are same as expected
// 2) There are no active tasks in the executor
// 3) Executor has shutdown correctly
ThreadPoolExecutor executor = getExecutor();
Assert.assertEquals(expectedCompletedTaskCount,
executor.getCompletedTaskCount());
Assert.assertEquals(0, executor.getActiveCount());
Assert.assertTrue(executor.isTerminated());
} else {
assert getExecutor() == null;
}
}
}
}

43
hadoop-common-project/hadoop-common/src/test/resources/testConf.xml
View File

@ -324,7 +324,7 @@
<comparators> <comparators>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^-cp \[-f\] \[-p \| -p\[topax\]\] \[-d\] &lt;src&gt; \.\.\. &lt;dst&gt; :\s*</expected-output> <expected-output>^-cp \[-f\] \[-p \| -p\[topax\]\] \[-d\] \[-t &lt;thread count&gt;\] \[-q &lt;thread pool queue size&gt;\] &lt;src&gt; \.\.\. &lt;dst&gt; :\s*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
@ -332,43 +332,64 @@
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^( |\t)*multiple files, the destination must be a directory.( )*Passing -p preserves status( )*</expected-output> <expected-output>^( |\t)*multiple files, the destination must be a directory.( )*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^( |\t)*\[topax\] \(timestamps, ownership, permission, ACLs, XAttr\). If -p is specified( )*</expected-output> <expected-output>^( |\t)*Flags :( )*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^( |\t)*with no &lt;arg&gt;, then preserves timestamps, ownership, permission. If -pa is( )*</expected-output> <expected-output>^( |\t)*-p\[topax\]\s+Preserve file attributes \[topx\] \(timestamps,( )*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^( |\t)*specified, then preserves permission also because ACL is a super-set of( )*</expected-output> <expected-output>^( |\t)*ownership, permission, ACL, XAttr\). If -p is( )*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^( |\t)*permission. Passing -f overwrites the destination if it already exists. raw( )*</expected-output> <expected-output>^( |\t)*specified with no arg, then preserves timestamps,( )*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^( |\t)*namespace extended attributes are preserved if \(1\) they are supported \(HDFS( )*</expected-output> <expected-output>^( |\t)*ownership, permission. If -pa is specified, then( )*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^( |\t)*only\) and, \(2\) all of the source and target pathnames are in the \/\.reserved\/raw( )*</expected-output> <expected-output>^( |\t)*preserves permission also because ACL is a( )*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^( |\t)*hierarchy. raw namespace xattr preservation is determined solely by the presence( )*</expected-output> <expected-output>^( |\t)*super-set of permission. Determination of whether( )*</expected-output>
</comparator>
<comparator>
<type>RegexpComparator</type>
<expected-output>^( |\t)*raw namespace extended attributes are preserved is( )*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^\s*\(or absence\) of the \/\.reserved\/raw prefix and not by the -p option\. Passing -d( )*</expected-output> <expected-output>^( |\t)*independent of the -p flag.( )*</expected-output>
</comparator> </comparator>
<comparator> <comparator>
<type>RegexpComparator</type> <type>RegexpComparator</type>
<expected-output>^\s*will skip creation of temporary file\(&lt;dst&gt;\._COPYING_\)\.( )*</expected-output> <expected-output>^\s*-f\s+Overwrite the destination if it already exists.( )*</expected-output>
</comparator>
<comparator>
<type>RegexpComparator</type>
<expected-output>^\s*-d\s+Skip creation of temporary file\(&lt;dst&gt;\._COPYING_\).( )*</expected-output>
</comparator>
<comparator>
<type>RegexpComparator</type>
<expected-output>^\s*-t &lt;thread count&gt;\s+Number of threads to be used, default is 1.( )*</expected-output>
</comparator>
<comparator>
<type>RegexpComparator</type>
<expected-output>^\s*-q &lt;thread pool queue size&gt;\s+Thread pool queue size to be used, default is( )*</expected-output>
</comparator>
<comparator>
<type>RegexpComparator</type>
<expected-output>^( |\t)*1024.\s*</expected-output>
</comparator> </comparator>
</comparators> </comparators>
</test> </test>