diff --git a/hadoop-common-project/hadoop-common/src/site/apt/DeprecatedProperties.apt.vm b/hadoop-common-project/hadoop-common/src/site/apt/DeprecatedProperties.apt.vm index 4d1b39d0dfd..32686d68e16 100644 --- a/hadoop-common-project/hadoop-common/src/site/apt/DeprecatedProperties.apt.vm +++ b/hadoop-common-project/hadoop-common/src/site/apt/DeprecatedProperties.apt.vm @@ -198,7 +198,9 @@ Deprecated Properties *---+---+ |mapred.compress.map.output | mapreduce.map.output.compress *---+---+ -|mapred.create.symlink | mapreduce.job.cache.symlink.create +|mapred.create.symlink | NONE - symlinking is always on +*---+---+ +|mapreduce.job.cache.symlink.create | NONE - symlinking is always on *---+---+ |mapred.data.field.separator | mapreduce.fieldsel.data.field.separator *---+---+ diff --git a/hadoop-mapreduce-project/CHANGES.txt b/hadoop-mapreduce-project/CHANGES.txt index d7297149244..89f67ebd549 100644 --- a/hadoop-mapreduce-project/CHANGES.txt +++ b/hadoop-mapreduce-project/CHANGES.txt @@ -647,6 +647,9 @@ Release 0.23.3 - UNRELEASED MAPREDUCE-4496. AM logs link is missing user name (Jason Lowe via bobby) + MAPREDUCE-4493. Distibuted Cache Compatability Issues (Robert Evans + via tgraves) + Release 0.23.2 - UNRELEASED INCOMPATIBLE CHANGES diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/main/java/org/apache/hadoop/mapred/LocalDistributedCacheManager.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/main/java/org/apache/hadoop/mapred/LocalDistributedCacheManager.java index 85cef6e2a6c..fb26245cd8e 100644 --- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/main/java/org/apache/hadoop/mapred/LocalDistributedCacheManager.java +++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/main/java/org/apache/hadoop/mapred/LocalDistributedCacheManager.java @@ -84,7 +84,6 @@ class LocalDistributedCacheManager { * @throws IOException */ public void setup(JobConf conf) throws IOException { - boolean mkLinks = DistributedCache.getSymlink(conf); File workDir = new File(System.getProperty("user.dir")); // Generate YARN local resources objects corresponding to the distributed @@ -145,11 +144,9 @@ class LocalDistributedCacheManager { throw new IOException(e); } String pathString = path.toUri().toString(); - if(mkLinks) { - String link = entry.getKey(); - String target = new File(path.toUri()).getPath(); - symlink(workDir, target, link); - } + String link = entry.getKey(); + String target = new File(path.toUri()).getPath(); + symlink(workDir, target, link); if (resource.getType() == LocalResourceType.ARCHIVE) { localArchives.add(pathString); diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapred/TestLocalDistributedCacheManager.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapred/TestLocalDistributedCacheManager.java index 368ac2bceda..b131473d63d 100644 --- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapred/TestLocalDistributedCacheManager.java +++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapred/TestLocalDistributedCacheManager.java @@ -150,7 +150,6 @@ public class TestLocalDistributedCacheManager { conf.set(MRJobConfig.CACHE_FILES_SIZES, "201"); conf.set(MRJobConfig.CACHE_FILE_VISIBILITIES, "false"); conf.set(MRConfig.LOCAL_DIR, localDir.getAbsolutePath()); - conf.set(MRJobConfig.CACHE_SYMLINK, "yes"); LocalDistributedCacheManager manager = new LocalDistributedCacheManager(); try { manager.setup(conf); @@ -197,7 +196,6 @@ public class TestLocalDistributedCacheManager { conf.set(MRJobConfig.CACHE_FILES, ""); conf.set(MRConfig.LOCAL_DIR, localDir.getAbsolutePath()); - conf.set(MRJobConfig.CACHE_SYMLINK, "yes"); LocalDistributedCacheManager manager = new LocalDistributedCacheManager(); try { manager.setup(conf); @@ -268,7 +266,6 @@ public class TestLocalDistributedCacheManager { conf.set(MRJobConfig.CACHE_FILES_SIZES, "201,201"); conf.set(MRJobConfig.CACHE_FILE_VISIBILITIES, "false,false"); conf.set(MRConfig.LOCAL_DIR, localDir.getAbsolutePath()); - conf.set(MRJobConfig.CACHE_SYMLINK, "yes"); LocalDistributedCacheManager manager = new LocalDistributedCacheManager(); try { manager.setup(conf); diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapred/TestMRWithDistributedCache.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapred/TestMRWithDistributedCache.java index c9ce7cba9b7..b704c77c96a 100644 --- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapred/TestMRWithDistributedCache.java +++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-common/src/test/java/org/apache/hadoop/mapred/TestMRWithDistributedCache.java @@ -146,7 +146,6 @@ public class TestMRWithDistributedCache extends TestCase { job.addFileToClassPath(second); job.addArchiveToClassPath(third); job.addCacheArchive(fourth.toUri()); - job.createSymlink(); job.setMaxMapAttempts(1); // speed up failures job.submit(); diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/filecache/DistributedCache.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/filecache/DistributedCache.java index 3b9424a81c9..0e7438fa4e6 100644 --- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/filecache/DistributedCache.java +++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/filecache/DistributedCache.java @@ -48,8 +48,12 @@ import org.apache.hadoop.mapreduce.Job; * Archives (zip, tar and tgz/tar.gz files) are un-archived at the slave nodes. * Jars may be optionally added to the classpath of the tasks, a rudimentary * software distribution mechanism. Files have execution permissions. - * Optionally users can also direct it to symlink the distributed cache file(s) - * into the working directory of the task.
+ * In older version of Hadoop Map/Reduce users could optionally ask for symlinks + * to be created in the working directory of the child task. In the current + * version symlinks are always created. If the URL does not have a fragment + * the name of the file or directory will be used. If multiple files or + * directories map to the same link name, the last one added, will be used. All + * others will not even be downloaded. * *DistributedCache
tracks modification timestamps of the cache
* files. Clearly the cache files should not be modified by the application
@@ -91,8 +95,7 @@ import org.apache.hadoop.mapreduce.Job;
*
* public void configure(JobConf job) {
* // Get the cached archives/files
- * localArchives = DistributedCache.getLocalCacheArchives(job);
- * localFiles = DistributedCache.getLocalCacheFiles(job);
+ * File f = new File("./map.zip/some/file/in/zip.txt");
* }
*
* public void map(K key, V value,
diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/mapred/pipes/Submitter.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/mapred/pipes/Submitter.java
index ddc3ae80172..57370872e30 100644
--- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/mapred/pipes/Submitter.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/mapred/pipes/Submitter.java
@@ -313,7 +313,6 @@ public class Submitter extends Configured implements Tool {
// add default debug script only when executable is expressed as
//
DistributedCache
tracks modification timestamps of the cache
* files. Clearly the cache files should not be modified by the application
@@ -98,8 +102,7 @@ import java.net.URI;
*
* public void configure(JobConf job) {
* // Get the cached archives/files
- * localArchives = DistributedCache.getLocalCacheArchives(job);
- * localFiles = DistributedCache.getLocalCacheFiles(job);
+ * File f = new File("./map.zip/some/file/in/zip.txt");
* }
*
* public void map(K key, V value,
@@ -375,32 +378,26 @@ public class DistributedCache {
}
/**
- * This method allows you to create symlinks in the current working directory
- * of the task to all the cache files/archives.
- * Intended to be used by user code.
+ * Originally intended to enable symlinks, but currently symlinks cannot be
+ * disabled. This is a NO-OP.
* @param conf the jobconf
- * @deprecated Use {@link Job#createSymlink()} instead
+ * @deprecated This is a NO-OP.
*/
@Deprecated
public static void createSymlink(Configuration conf){
- conf.set(MRJobConfig.CACHE_SYMLINK, "yes");
+ //NOOP
}
/**
- * This method checks to see if symlinks are to be create for the
- * localized cache files in the current working directory
- * Used by internal DistributedCache code.
+ * Originally intended to check if symlinks should be used, but currently
+ * symlinks cannot be disabled.
* @param conf the jobconf
- * @return true if symlinks are to be created- else return false
- * @deprecated Use {@link JobContext#getSymlink()} instead
+ * @return true
+ * @deprecated symlinks are always created.
*/
@Deprecated
public static boolean getSymlink(Configuration conf){
- String result = conf.get(MRJobConfig.CACHE_SYMLINK);
- if ("yes".equals(result)){
- return true;
- }
- return false;
+ return true;
}
private static boolean[] parseBooleans(String[] strs) {
diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/mapreduce/util/ConfigUtil.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/mapreduce/util/ConfigUtil.java
index fde6aa4647f..f5dbc3a7f1a 100644
--- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/mapreduce/util/ConfigUtil.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core/src/main/java/org/apache/hadoop/mapreduce/util/ConfigUtil.java
@@ -246,8 +246,6 @@ public class ConfigUtil {
new String[] {MRJobConfig.CACHE_FILE_TIMESTAMPS});
Configuration.addDeprecation("mapred.cache.archives.timestamps",
new String[] {MRJobConfig.CACHE_ARCHIVES_TIMESTAMPS});
- Configuration.addDeprecation("mapred.create.symlink",
- new String[] {MRJobConfig.CACHE_SYMLINK});
Configuration.addDeprecation("mapred.working.dir",
new String[] {MRJobConfig.WORKING_DIR});
Configuration.addDeprecation("user.name",
diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapred/MRCaching.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapred/MRCaching.java
index 02b6e58d0d2..2d5b8ccb096 100644
--- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapred/MRCaching.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapred/MRCaching.java
@@ -210,19 +210,10 @@ public class MRCaching {
fs.copyFromLocalFile(tarPath1, cachePath);
fs.copyFromLocalFile(tarPath2, cachePath);
}
-
- public static TestResult launchMRCache(String indir,
- String outdir, String cacheDir,
- JobConf conf, String input)
- throws IOException {
- setupCache(cacheDir, FileSystem.get(conf));
- return launchMRCache(indir,outdir, cacheDir, conf, input, false);
- }
public static TestResult launchMRCache(String indir,
String outdir, String cacheDir,
- JobConf conf, String input,
- boolean withSymlink)
+ JobConf conf, String input)
throws IOException {
String TEST_ROOT_DIR = new Path(System.getProperty("test.build.data","/tmp"))
.toString().replace(' ', '+');
@@ -256,24 +247,13 @@ public class MRCaching {
conf.setNumReduceTasks(1);
conf.setSpeculativeExecution(false);
URI[] uris = new URI[6];
- if (!withSymlink) {
- conf.setMapperClass(MRCaching.MapClass.class);
- uris[0] = fs.getUri().resolve(cacheDir + "/test.txt");
- uris[1] = fs.getUri().resolve(cacheDir + "/test.jar");
- uris[2] = fs.getUri().resolve(cacheDir + "/test.zip");
- uris[3] = fs.getUri().resolve(cacheDir + "/test.tgz");
- uris[4] = fs.getUri().resolve(cacheDir + "/test.tar.gz");
- uris[5] = fs.getUri().resolve(cacheDir + "/test.tar");
- } else {
- DistributedCache.createSymlink(conf);
- conf.setMapperClass(MRCaching.MapClass2.class);
- uris[0] = fs.getUri().resolve(cacheDir + "/test.txt#" + "test.txt");
- uris[1] = fs.getUri().resolve(cacheDir + "/test.jar#" + "testjar");
- uris[2] = fs.getUri().resolve(cacheDir + "/test.zip#" + "testzip");
- uris[3] = fs.getUri().resolve(cacheDir + "/test.tgz#" + "testtgz");
- uris[4] = fs.getUri().resolve(cacheDir + "/test.tar.gz#" + "testtargz");
- uris[5] = fs.getUri().resolve(cacheDir + "/test.tar#" + "testtar");
- }
+ conf.setMapperClass(MRCaching.MapClass2.class);
+ uris[0] = fs.getUri().resolve(cacheDir + "/test.txt");
+ uris[1] = fs.getUri().resolve(cacheDir + "/test.jar");
+ uris[2] = fs.getUri().resolve(cacheDir + "/test.zip");
+ uris[3] = fs.getUri().resolve(cacheDir + "/test.tgz");
+ uris[4] = fs.getUri().resolve(cacheDir + "/test.tar.gz");
+ uris[5] = fs.getUri().resolve(cacheDir + "/test.tar");
DistributedCache.addCacheFile(uris[0], conf);
// Save expected file sizes
diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapred/TestMiniMRDFSCaching.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapred/TestMiniMRDFSCaching.java
index 6e8abd73a06..cda200e97e5 100644
--- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapred/TestMiniMRDFSCaching.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapred/TestMiniMRDFSCaching.java
@@ -48,7 +48,7 @@ public class TestMiniMRDFSCaching extends TestCase {
"/cachedir",
mr.createJobConf(),
"The quick brown fox\nhas many silly\n"
- + "red fox sox\n", false);
+ + "red fox sox\n");
assertTrue("Archives not matching", ret.isOutputOk);
// launch MR cache with symlinks
ret = MRCaching.launchMRCache("/testing/wc/input",
@@ -56,7 +56,7 @@ public class TestMiniMRDFSCaching extends TestCase {
"/cachedir",
mr.createJobConf(),
"The quick brown fox\nhas many silly\n"
- + "red fox sox\n", true);
+ + "red fox sox\n");
assertTrue("Archives not matching", ret.isOutputOk);
} finally {
if (fileSys != null) {
diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapreduce/v2/TestMRJobs.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapreduce/v2/TestMRJobs.java
index 59b8689b20d..7b6ecd25560 100644
--- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapreduce/v2/TestMRJobs.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapreduce/v2/TestMRJobs.java
@@ -211,6 +211,7 @@ public class TestMRJobs {
Path outputDir =
new Path(mrCluster.getTestWorkDir().getAbsolutePath(), "random-output");
FileOutputFormat.setOutputPath(job, outputDir);
+ job.setSpeculativeExecution(false);
job.addFileToClassPath(APP_JAR); // The AppMaster jar itself.
job.setJarByClass(RandomTextWriterJob.class);
job.setMaxMapAttempts(1); // speed up failures
@@ -462,7 +463,6 @@ public class TestMRJobs {
job.addFileToClassPath(APP_JAR); // The AppMaster jar itself.
job.addArchiveToClassPath(third);
job.addCacheArchive(fourth.toUri());
- job.createSymlink();
job.setMaxMapAttempts(1); // speed up failures
job.submit();
diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapreduce/v2/TestSpeculativeExecution.java b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapreduce/v2/TestSpeculativeExecution.java
index 90f7cafc632..17d82d6dacc 100644
--- a/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapreduce/v2/TestSpeculativeExecution.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-jobclient/src/test/java/org/apache/hadoop/mapreduce/v2/TestSpeculativeExecution.java
@@ -301,7 +301,6 @@ public class TestSpeculativeExecution {
// Creates the Job Configuration
job.addFileToClassPath(APP_JAR); // The AppMaster jar itself.
- job.createSymlink();
job.setMaxMapAttempts(2);
job.submit();
diff --git a/hadoop-mapreduce-project/hadoop-mapreduce-examples/src/main/java/org/apache/hadoop/examples/Sort.java b/hadoop-mapreduce-project/hadoop-mapreduce-examples/src/main/java/org/apache/hadoop/examples/Sort.java
index 901934bbf98..5184bf0a58a 100644
--- a/hadoop-mapreduce-project/hadoop-mapreduce-examples/src/main/java/org/apache/hadoop/examples/Sort.java
+++ b/hadoop-mapreduce-project/hadoop-mapreduce-examples/src/main/java/org/apache/hadoop/examples/Sort.java
@@ -167,7 +167,6 @@ public class Sort