Author: stevel
Date: Tue Jun 9 16:11:19 2009
New Revision: 783055
URL: http://svn.apache.org/viewvc?rev=783055&view=rev
Log:
sync up with SVN HEAD
Added:
hadoop/core/branches/HADOOP-3628-2/lib/jsp-2.1/
- copied from r779668, hadoop/core/trunk/lib/jsp-2.1/
hadoop/core/branches/HADOOP-3628-2/lib/jsp-2.1/jsp-2.1.jar
- copied unchanged from r779668, hadoop/core/trunk/lib/jsp-2.1/jsp-2.1.jar
hadoop/core/branches/HADOOP-3628-2/lib/jsp-2.1/jsp-api-2.1.jar
- copied unchanged from r779668,
hadoop/core/trunk/lib/jsp-2.1/jsp-api-2.1.jar
hadoop/core/branches/HADOOP-3628-2/src/contrib/sqoop/src/java/org/apache/hadoop/sqoop/manager/LocalMySQLManager.java
- copied unchanged from r782987,
hadoop/core/trunk/src/contrib/sqoop/src/java/org/apache/hadoop/sqoop/manager/LocalMySQLManager.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/sqoop/src/test/org/apache/hadoop/sqoop/manager/LocalMySQLTest.java
- copied unchanged from r782987,
hadoop/core/trunk/src/contrib/sqoop/src/test/org/apache/hadoop/sqoop/manager/LocalMySQLTest.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/streaming/src/test/org/apache/hadoop/streaming/TestStreamingStatus.java
- copied unchanged from r779668,
hadoop/core/trunk/src/contrib/streaming/src/test/org/apache/hadoop/streaming/TestStreamingStatus.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapreduce/lib/input/CombineFileInputFormat.java
- copied unchanged from r780635,
hadoop/core/trunk/src/mapred/org/apache/hadoop/mapreduce/lib/input/CombineFileInputFormat.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapreduce/lib/input/CombineFileRecordReader.java
- copied unchanged from r780635,
hadoop/core/trunk/src/mapred/org/apache/hadoop/mapreduce/lib/input/CombineFileRecordReader.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapreduce/lib/input/CombineFileSplit.java
- copied unchanged from r780635,
hadoop/core/trunk/src/mapred/org/apache/hadoop/mapreduce/lib/input/CombineFileSplit.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestMapredHeartbeat.java
- copied unchanged from r782987,
hadoop/core/trunk/src/test/mapred/org/apache/hadoop/mapred/TestMapredHeartbeat.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestMiniMRChildTask.java
- copied unchanged from r782987,
hadoop/core/trunk/src/test/mapred/org/apache/hadoop/mapred/TestMiniMRChildTask.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestRunningTaskLimits.java
- copied unchanged from r782987,
hadoop/core/trunk/src/test/mapred/org/apache/hadoop/mapred/TestRunningTaskLimits.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapreduce/lib/input/
- copied from r780635,
hadoop/core/trunk/src/test/mapred/org/apache/hadoop/mapreduce/lib/input/
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapreduce/lib/input/TestCombineFileInputFormat.java
- copied unchanged from r780635,
hadoop/core/trunk/src/test/mapred/org/apache/hadoop/mapreduce/lib/input/TestCombineFileInputFormat.java
Removed:
hadoop/core/branches/HADOOP-3628-2/lib/commons-cli-2.0-SNAPSHOT.jar
hadoop/core/branches/HADOOP-3628-2/src/test/hdfs/org/apache/hadoop/hdfs/server/namenode/TestINode.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestMiniMRTaskTempDir.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/lib/TestCombineFileInputFormat.java
Modified:
hadoop/core/branches/HADOOP-3628-2/ (props changed)
hadoop/core/branches/HADOOP-3628-2/.eclipse.templates/.classpath
hadoop/core/branches/HADOOP-3628-2/CHANGES.txt (contents, props changed)
hadoop/core/branches/HADOOP-3628-2/ivy.xml
hadoop/core/branches/HADOOP-3628-2/ivy/libraries.properties
hadoop/core/branches/HADOOP-3628-2/src/contrib/build-contrib.xml
hadoop/core/branches/HADOOP-3628-2/src/contrib/build.xml
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/ivy.xml
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacitySchedulerConf.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacityTaskScheduler.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/MemoryMatcher.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/test/org/apache/hadoop/mapred/TestCapacityScheduler.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/eclipse-plugin/build.xml
hadoop/core/branches/HADOOP-3628-2/src/contrib/hdfsproxy/bin/hdfsproxy
hadoop/core/branches/HADOOP-3628-2/src/contrib/hdfsproxy/build.xml
hadoop/core/branches/HADOOP-3628-2/src/contrib/hdfsproxy/ivy.xml
hadoop/core/branches/HADOOP-3628-2/src/contrib/sqoop/src/java/org/apache/hadoop/sqoop/ConnFactory.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/sqoop/src/java/org/apache/hadoop/sqoop/ImportOptions.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/sqoop/src/java/org/apache/hadoop/sqoop/manager/MySQLManager.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/sqoop/src/java/org/apache/hadoop/sqoop/mapred/ImportJob.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/sqoop/src/java/org/apache/hadoop/sqoop/orm/CompilationManager.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/sqoop/src/test/org/apache/hadoop/sqoop/AllTests.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/streaming/ivy.xml
hadoop/core/branches/HADOOP-3628-2/src/contrib/streaming/src/java/org/apache/hadoop/streaming/PipeMapRed.java
hadoop/core/branches/HADOOP-3628-2/src/contrib/streaming/src/test/org/apache/hadoop/streaming/StderrApp.java
hadoop/core/branches/HADOOP-3628-2/src/core/core-default.xml
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/filecache/DistributedCache.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/fs/FsShell.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/fs/kfs/KosmosFileSystem.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/fs/s3/S3FileSystem.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/fs/s3native/Jets3tNativeFileSystemStore.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/fs/s3native/NativeFileSystemStore.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/fs/s3native/NativeS3FileSystem.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/io/IOUtils.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/metrics/ContextFactory.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/util/GenericOptionsParser.java
hadoop/core/branches/HADOOP-3628-2/src/core/org/apache/hadoop/util/ProcfsBasedProcessTree.java
hadoop/core/branches/HADOOP-3628-2/src/docs/src/documentation/content/xdocs/hdfs_design.xml
hadoop/core/branches/HADOOP-3628-2/src/docs/src/documentation/content/xdocs/hdfs_user_guide.xml
hadoop/core/branches/HADOOP-3628-2/src/examples/org/apache/hadoop/examples/MultiFileWordCount.java
hadoop/core/branches/HADOOP-3628-2/src/examples/org/apache/hadoop/examples/Sort.java
hadoop/core/branches/HADOOP-3628-2/src/examples/org/apache/hadoop/examples/dancing/DistributedPentomino.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/DFSClient.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/common/StorageInfo.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/datanode/DataBlockScanner.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/datanode/DataNode.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/datanode/DataStorage.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/datanode/DatanodeBlockInfo.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/namenode/BlockManager.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/namenode/FSEditLog.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/namenode/FSNamesystem.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/namenode/INode.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/namenode/INodeDirectory.java
hadoop/core/branches/HADOOP-3628-2/src/hdfs/org/apache/hadoop/hdfs/server/namenode/SecondaryNameNode.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/mapred-default.xml
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/JobClient.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/JobConf.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/JobHistory.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/JobInProgress.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/JobTracker.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/MRConstants.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/MapTask.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/Merger.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/ReduceTask.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/ResourceEstimator.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/TaskMemoryManagerThread.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/TaskRunner.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/TaskTracker.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/lib/CombineFileInputFormat.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/lib/CombineFileRecordReader.java
hadoop/core/branches/HADOOP-3628-2/src/mapred/org/apache/hadoop/mapred/lib/CombineFileSplit.java
hadoop/core/branches/HADOOP-3628-2/src/test/bin/test-patch.sh
hadoop/core/branches/HADOOP-3628-2/src/test/core/org/apache/hadoop/fs/s3/S3FileSystemContractBaseTest.java
hadoop/core/branches/HADOOP-3628-2/src/test/core/org/apache/hadoop/fs/s3native/InMemoryNativeFileSystemStore.java
hadoop/core/branches/HADOOP-3628-2/src/test/core/org/apache/hadoop/fs/s3native/NativeS3FileSystemContractBaseTest.java
hadoop/core/branches/HADOOP-3628-2/src/test/core/org/apache/hadoop/util/TestProcfsBasedProcessTree.java
hadoop/core/branches/HADOOP-3628-2/src/test/findbugsExcludeFile.xml
hadoop/core/branches/HADOOP-3628-2/src/test/hdfs-with-mr/org/apache/hadoop/fs/TestCopyFiles.java
hadoop/core/branches/HADOOP-3628-2/src/test/hdfs-with-mr/org/apache/hadoop/hdfs/NNBench.java
hadoop/core/branches/HADOOP-3628-2/src/test/hdfs/org/apache/hadoop/hdfs/TestDFSShellGenericOptions.java
hadoop/core/branches/HADOOP-3628-2/src/test/hdfs/org/apache/hadoop/hdfs/TestDFSStartupVersions.java
hadoop/core/branches/HADOOP-3628-2/src/test/hdfs/org/apache/hadoop/hdfs/TestFileCreation.java
hadoop/core/branches/HADOOP-3628-2/src/test/hdfs/org/apache/hadoop/hdfs/TestFileStatus.java
hadoop/core/branches/HADOOP-3628-2/src/test/hdfs/org/apache/hadoop/hdfs/TestSeekBug.java
hadoop/core/branches/HADOOP-3628-2/src/test/hdfs/org/apache/hadoop/hdfs/TestSmallBlock.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/SortValidator.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestEmptyJob.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestFileOutputFormat.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestJavaSerialization.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestJobTrackerRestart.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestMiniMRDFSSort.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/TestTaskTrackerMemoryManager.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapred/UtilsForTests.java
hadoop/core/branches/HADOOP-3628-2/src/test/mapred/org/apache/hadoop/mapreduce/TestMapReduceLocal.java
hadoop/core/branches/HADOOP-3628-2/src/tools/org/apache/hadoop/tools/DistCp.java
hadoop/core/branches/HADOOP-3628-2/src/webapps/job/jobdetails.jsp
Propchange: hadoop/core/branches/HADOOP-3628-2/
------------------------------------------------------------------------------
--- svn:mergeinfo (original)
+++ svn:mergeinfo Tue Jun 9 16:11:19 2009
@@ -1 +1,2 @@
/hadoop/core/branches/branch-0.19:713112
+/hadoop/core/trunk:779103-783032
Modified: hadoop/core/branches/HADOOP-3628-2/.eclipse.templates/.classpath
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/.eclipse.templates/.classpath?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
--- hadoop/core/branches/HADOOP-3628-2/.eclipse.templates/.classpath (original)
+++ hadoop/core/branches/HADOOP-3628-2/.eclipse.templates/.classpath Tue Jun 9
16:11:19 2009
@@ -14,11 +14,10 @@
<classpathentry kind="src" path="src/contrib/streaming/src/test"/>
<classpathentry kind="con"
path="org.eclipse.jdt.launching.JRE_CONTAINER"/>
<classpathentry kind="var" path="ANT_HOME/lib/ant.jar"/>
- <classpathentry kind="lib" path="lib/commons-cli-2.0-SNAPSHOT.jar"/>
<classpathentry kind="lib" path="lib/hsqldb-1.8.0.10.jar"/>
<classpathentry kind="lib" path="lib/kfs-0.2.2.jar"/>
- <classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/jsp-2.1-6.1.14.jar"/>
- <classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/jsp-api-2.1-6.1.14.jar"/>
+ <classpathentry kind="lib" path="lib/jsp-2.1/jsp-2.1.jar"/>
+ <classpathentry kind="lib" path="lib/jsp-2.1/jsp-api-2.1.jar"/>
<classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/commons-codec-1.3.jar"/>
<classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/commons-httpclient-3.0.1.jar"/>
<classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/commons-el-1.0.jar"/>
@@ -38,6 +37,8 @@
<classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/slf4j-api-1.4.3.jar"/>
<classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/slf4j-log4j12-1.4.3.jar"/>
<classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/xmlenc-0.52.jar"/>
+ <classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/commons-cli-1.2.jar"/>
+ <classpathentry kind="lib"
path="build/ivy/lib/Hadoop/common/commons-cli-2.0-mahout.jar"/>
<classpathentry kind="lib"
path="src/test/lib/ftplet-api-1.0.0-SNAPSHOT.jar"/>
<classpathentry kind="lib"
path="src/test/lib/ftpserver-core-1.0.0-SNAPSHOT.jar"/>
<classpathentry kind="lib"
path="src/test/lib/ftpserver-server-1.0.0-SNAPSHOT.jar"/>
Modified: hadoop/core/branches/HADOOP-3628-2/CHANGES.txt
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/CHANGES.txt?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
--- hadoop/core/branches/HADOOP-3628-2/CHANGES.txt (original)
+++ hadoop/core/branches/HADOOP-3628-2/CHANGES.txt Tue Jun 9 16:11:19 2009
@@ -131,6 +131,12 @@
HADOOP-4861. Add disk usage with human-readable size (-duh).
(Todd Lipcon via tomwhite)
+ HADOOP-5844. Use mysqldump when connecting to local mysql instance in
Sqoop.
+ (Aaron Kimball via tomwhite)
+
+ HADOOP-5170. Allows jobs to set max maps/reduces per-node and per-cluster.
+ (Matei Zaharia via ddas)
+
IMPROVEMENTS
HADOOP-4565. Added CombineFileInputFormat to use data locality information
@@ -349,8 +355,6 @@
HADOOP-5780. Improve per block message prited by "-metaSave" in HDFS.
(Raghu Angadi)
-
- HADOOP-5792. To resolve jsp-2.1 jars through ivy (Giridharan Kesavan)
HADOOP-5823. Added a new class DeprecatedUTF8 to help with removing
UTF8 related javac warnings. These warnings are removed in
@@ -392,12 +396,48 @@
TestNameEditsConfig, TestStartup and TestStorageRestore.
(Jakob Homan via shv)
- HADOOP-5438. Provide a single FileSystem method to create or
open-for-append
+ HADOOP-5438. Provide a single FileSystem method to create or
open-for-append
to a file. (He Yongqiang via dhruba)
HADOOP-5472. Change DistCp to support globbing of input paths. (Dhruba
Borthakur and Rodrigo Schmidt via szetszwo)
+ HADOOP-5175. Don't unpack libjars on classpath. (Todd Lipcon via tomwhite)
+
+ HADOOP-5620. Add an option to DistCp for preserving modification and access
+ times. (Rodrigo Schmidt via szetszwo)
+
+ HADOOP-5664. Change map serialization so a lock is obtained only where
+ contention is possible, rather than for each write. (cdouglas)
+
+ HADOOP-5896. Remove the dependency of GenericOptionsParser on
+ Option.withArgPattern. (Giridharan Kesavan and Sharad Agarwal via
+ sharad)
+
+ HADOOP-5698. Change org.apache.hadoop.examples.MultiFileWordCount to
+ use new mapreduce api. (Amareshwari Sriramadasu via sharad)
+
+ HADOOP-5696. Change org.apache.hadoop.examples.Sort to use new
+ mapreduce api. (Amareshwari Sriramadasu via sharad)
+
+ HADOOP-5784. Makes the number of heartbeats that should arrive a second
+ at the JobTracker configurable. (Amareshwari Sriramadasu via ddas)
+
+ HADOOP-5955. Changes TestFileOuputFormat so that is uses LOCAL_MR
+ instead of CLUSTER_MR. (Jothi Padmanabhan via das)
+
+ HADOOP-5948. Changes TestJavaSerialization to use LocalJobRunner
+ instead of MiniMR/DFS cluster. (Jothi Padmanabhan via das)
+
+ HADOOP-2838. Add mapred.child.env to pass environment variables to
+ tasktracker's child processes. (Amar Kamat via sharad)
+
+ HADOOP-5961. DataNode process understand generic hadoop command line
+ options (like -Ddfs.property=value). (Raghu Angadi)
+
+ HADOOP-5694. Change org.apache.hadoop.examples.dancing to use new
+ mapreduce api. (Amareshwari Sriramadasu via sharad)
+
OPTIMIZATIONS
HADOOP-5595. NameNode does not need to run a replicator to choose a
@@ -695,8 +735,68 @@
HADOOP-5635. Change distributed cache to work with other distributed file
systems. (Andrew Hitchcock via tomwhite)
- HADOOP-5700. INode.getPathComponents throws NPE when given a non-absolute
- path. (Todd Lipcon via tomwhite)
+ HADOOP-5856. Fix "unsafe multithreaded use of DateFormat" findbugs warning
+ in DataBlockScanner. (Kan Zhang via szetszwo)
+
+ HADOOP-4864. Fixes a problem to do with -libjars with multiple jars when
+ client and cluster reside on different OSs. (Amareshwari Sriramadasu via
ddas)
+
+ HADOOP-5623. Fixes a problem to do with status messages getting overwritten
+ in streaming jobs. (Rick Cox and Jothi Padmanabhan via ddas)
+
+ HADOOP-5895. Fixes computation of count of merged bytes for logging.
+ (Ravi Gummadi via ddas)
+
+ HADOOP-5805. problem using top level s3 buckets as input/output
directories.
+ (Ian Nowland via tomwhite)
+
+ HADOOP-5940. trunk eclipse-plugin build fails while trying to copy
+ commons-cli jar from the lib dir (Giridharan Kesavan via gkesavan)
+
+ HADOOP-5864. Fix DMI and OBL findbugs in packages hdfs and metrics.
+ (hairong)
+
+ HADOOP-5935. Fix Hudson's release audit warnings link is broken.
+ (Giridharan Kesavan via gkesavan)
+
+ HADOOP-5947. Delete empty TestCombineFileInputFormat.java
+
+ HADOOP-5899. Move a log message in FSEditLog to the right place for
+ avoiding unnecessary log. (Suresh Srinivas via szetszwo)
+
+ HADOOP-5944. Add Apache license header to BlockManager.java. (Suresh
+ Srinivas via szetszwo)
+
+ HADOOP-5891. SecondaryNamenode is able to converse with the NameNode
+ even when the default value of dfs.http.address is not overridden.
+ (Todd Lipcon via dhruba)
+
+ HADOOP-5953. The isDirectory(..) and isFile(..) methods in KosmosFileSystem
+ should not be deprecated. (szetszwo)
+
+ HADOOP-5954. Fix javac warnings in TestFileCreation, TestSmallBlock,
+ TestFileStatus, TestDFSShellGenericOptions, TestSeekBug and
+ TestDFSStartupVersions. (szetszwo)
+
+ HADOOP-5956. Fix ivy dependency in hdfsproxy and capacity-scheduler.
+ (Giridharan Kesavan via szetszwo)
+
+ HADOOP-5836. Bug in S3N handling of directory markers using an object with
+ a trailing "/" causes jobs to fail. (Ian Nowland via tomwhite)
+
+ HADOOP-5861. s3n files are not getting split by default. (tomwhite)
+
+ HADOOP-5762. Fix a problem that DistCp does not copy empty directory.
+ (Rodrigo Schmidt via szetszwo)
+
+ HADOOP-5859. Fix "wait() or sleep() with locks held" findbugs warnings in
+ DFSClient. (Kan Zhang via szetszwo)
+
+ HADOOP-5457. Fix to continue to run builds even if contrib test fails
+ (Giridharan Kesavan via gkesavan)
+
+ HADOOP-5963. Remove an unnecessary exception catch in NNBench. (Boris
+ Shkolnik via szetszwo)
Release 0.20.1 - Unreleased
@@ -780,6 +880,42 @@
HADOOP-5850. Fixes a problem to do with not being able to jobs with
0 maps/reduces. (Vinod K V via ddas)
+ HADOOP-4626. Correct the API links in hdfs forrest doc so that they
+ point to the same version of hadoop. (szetszwo)
+
+ HADOOP-5883. Fixed tasktracker memory monitoring to account for
+ momentary spurts in memory usage due to java's fork() model.
+ (yhemanth)
+
+ HADOOP-5539. Fixes a problem to do with not preserving intermediate
+ output compression for merged data.
+ (Jothi Padmanabhan and Billy Pearson via ddas)
+
+ HADOOP-5932. Fixes a problem in capacity scheduler in computing
+ available memory on a tasktracker.
+ (Vinod Kumar Vavilapalli via yhemanth)
+
+ HADOOP-5937. Correct a safemode message in FSNamesystem. (Ravi Phulari
+ via szetszwo)
+
+ HADOOP-5908. Fixes a problem to do with ArithmeticException in the
+ JobTracker when there are jobs with 0 maps. (Amar Kamat via ddas)
+
+ HADOOP-5924. Fixes a corner case problem to do with job recovery with
+ empty history files. Also, after a JT restart, sends KillTaskAction to
+ tasks that report back but the corresponding job hasn't been initialized
+ yet. (Amar Kamat via ddas)
+
+ HADOOP-5882. Fixes a reducer progress update problem for new mapreduce
+ api. (Amareshwari Sriramadasu via sharad)
+
+ HADOOP-5746. Fixes a corner case problem in Streaming, where if an
exception
+ happens in MROutputThread after the last call to the map/reduce method,
the
+ exception goes undetected. (Amar Kamat via ddas)
+
+ HADOOP-5884. Fixes accounting in capacity scheduler so that high RAM jobs
+ take more slots. (Vinod Kumar Vavilapalli via yhemanth)
+
Release 0.20.0 - 2009-04-15
INCOMPATIBLE CHANGES
@@ -1544,7 +1680,7 @@
(Rahul Kumar Singh via yhemanth)
HADOOP-5328. Fixes a problem in the renaming of job history files during
- job recovery. Amar Kamat via ddas)
+ job recovery. (Amar Kamat via ddas)
HADOOP-5417. Don't ignore InterruptedExceptions that happen when calling
into rpc. (omalley)
@@ -1657,7 +1793,7 @@
(ddas)
HADOOP-5392. Fixes a problem to do with JT crashing during recovery when
- the job files are garbled. (Amar Kamat vi ddas)
+ the job files are garbled. (Amar Kamat via ddas)
HADOOP-5332. Appending to files is not allowed (by default) unless
dfs.support.append is set to true. (dhruba)
@@ -1718,6 +1854,9 @@
HADOOP-5816. Fixes a problem in the KeyFieldBasedComparator to do with
ArrayIndexOutOfBounds exception. (He Yongqiang via ddas)
+ HADOOP-5951. Add Apache license header to StorageInfo.java. (Suresh
+ Srinivas via szetszwo)
+
Release 0.19.1 - 2009-02-23
IMPROVEMENTS
@@ -2361,7 +2500,7 @@
(Lohit Vijayarenu via mahadev)
HADOOP-3778. DFSInputStream.seek() did not retry in case of some errors.
- (LN via rangadi)
+ (Luo Ning via rangadi)
HADOOP-3661. The handling of moving files deleted through fuse-dfs to
Trash made similar to the behaviour from dfs shell.
Propchange: hadoop/core/branches/HADOOP-3628-2/CHANGES.txt
------------------------------------------------------------------------------
--- svn:mergeinfo (original)
+++ svn:mergeinfo Tue Jun 9 16:11:19 2009
@@ -1,2 +1,3 @@
/hadoop/core/branches/branch-0.18/CHANGES.txt:727226
/hadoop/core/branches/branch-0.19/CHANGES.txt:713112
+/hadoop/core/trunk/CHANGES.txt:779103-783032
Modified: hadoop/core/branches/HADOOP-3628-2/ivy.xml
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/ivy.xml?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
--- hadoop/core/branches/HADOOP-3628-2/ivy.xml (original)
+++ hadoop/core/branches/HADOOP-3628-2/ivy.xml Tue Jun 9 16:11:19 2009
@@ -78,13 +78,14 @@
<artifact conf="master"/>
</publications>
<dependencies>
-
- <!--used client side-->
-<!-- <dependency org="commons-cli"
- name="commons-cli"
- rev="${commons-cli.version}"
- conf="client->default"/> -->
-
+ <dependency org="org.apache.mahout.commons"
+ name="commons-cli"
+ rev="${commons-cli2.version}"
+ conf="common->default"/>
+ <dependency org="commons-cli"
+ name="commons-cli"
+ rev="${commons-cli.version}"
+ conf="common->default"/>
<dependency org="checkstyle"
name="checkstyle"
rev="${checkstyle.version}"
@@ -98,7 +99,6 @@
rev="${xerces.version}"
conf="jdiff->default">
</dependency>
-
<dependency org="xmlenc"
name="xmlenc"
rev="${xmlenc.version}"
@@ -149,14 +149,11 @@
name="jasper-compiler"
rev="${jasper.version}"
conf="jetty->master"/>
- <dependency org="org.mortbay.jetty"
- name="jsp-api-2.1"
- rev="${jetty.version}"
- conf="jetty->master"/>
- <dependency org="org.mortbay.jetty"
- name="jsp-2.1"
- rev="${jetty.version}"
- conf="jetty->master"/>
+<!-- this is resolved locally from the lib folder
+ <dependency org="tomcat"
+ name="jsp-api"
+ rev="${jsp-api.version}"
+ conf="jetty->master"/> -->
<dependency org="commons-el"
name="commons-el"
rev="${commons-el.version}"
Modified: hadoop/core/branches/HADOOP-3628-2/ivy/libraries.properties
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/ivy/libraries.properties?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
--- hadoop/core/branches/HADOOP-3628-2/ivy/libraries.properties (original)
+++ hadoop/core/branches/HADOOP-3628-2/ivy/libraries.properties Tue Jun 9
16:11:19 2009
@@ -18,7 +18,8 @@
checkstyle.version=4.2
-commons-cli.version=2.0-SNAPSHOT
+commons-cli.version=1.2
+commons-cli2.version=2.0-mahout
commons-codec.version=1.3
commons-collections.version=3.1
commons-httpclient.version=3.0.1
Modified: hadoop/core/branches/HADOOP-3628-2/src/contrib/build-contrib.xml
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/src/contrib/build-contrib.xml?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
--- hadoop/core/branches/HADOOP-3628-2/src/contrib/build-contrib.xml (original)
+++ hadoop/core/branches/HADOOP-3628-2/src/contrib/build-contrib.xml Tue Jun 9
16:11:19 2009
@@ -41,6 +41,7 @@
<property name="test.junit.output.format" value="plain"/>
<property name="test.output" value="no"/>
<property name="test.timeout" value="900000"/>
+ <property name="build.contrib.dir" location="${hadoop.root}/build/contrib"/>
<property name="build.dir" location="${hadoop.root}/build/contrib/${name}"/>
<property name="build.classes" location="${build.dir}/classes"/>
<property name="build.test" location="${build.dir}/test"/>
@@ -247,7 +248,12 @@
<fileset dir="${src.test}" includes="**/${testcase}.java"/>
</batchtest>
</junit>
- <fail if="tests.failed">Tests failed!</fail>
+ <antcall target="checkfailure"/>
+ </target>
+
+ <target name="checkfailure" if="tests.failed">
+ <touch file="${build.contrib.dir}/testsfailed"/>
+ <fail unless="continueOnFailure">Contrib Tests failed!</fail>
</target>
<!-- ================================================================== -->
Modified: hadoop/core/branches/HADOOP-3628-2/src/contrib/build.xml
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/src/contrib/build.xml?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
--- hadoop/core/branches/HADOOP-3628-2/src/contrib/build.xml (original)
+++ hadoop/core/branches/HADOOP-3628-2/src/contrib/build.xml Tue Jun 9
16:11:19 2009
@@ -45,16 +45,21 @@
<!-- Test all the contribs. -->
<!-- ====================================================== -->
<target name="test">
+ <property name="hadoop.root" location="${root}/../../../"/>
+ <property name="build.contrib.dir"
location="${hadoop.root}/build/contrib"/>
+ <delete file="${build.contrib.dir}/testsfailed"/>
<subant target="test">
+ <property name="continueOnFailure" value="true"/>
<fileset dir="." includes="hdfsproxy/build.xml"/>
<fileset dir="." includes="streaming/build.xml"/>
<fileset dir="." includes="fairscheduler/build.xml"/>
<fileset dir="." includes="capacity-scheduler/build.xml"/>
<fileset dir="." includes="mrunit/build.xml"/>
</subant>
+ <available file="${build.contrib.dir}/testsfailed" property="testsfailed"/>
+ <fail if="testsfailed">Tests failed!</fail>
</target>
-
<!-- ====================================================== -->
<!-- Clean all the contribs. -->
<!-- ====================================================== -->
Modified:
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/ivy.xml
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/ivy.xml?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
--- hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/ivy.xml
(original)
+++ hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/ivy.xml
Tue Jun 9 16:11:19 2009
@@ -24,6 +24,10 @@
<artifact conf="master"/>
</publications>
<dependencies>
+ <dependency org="commons-cli"
+ name="commons-cli"
+ rev="${commons-cli.version}"
+ conf="common->default"/>
<dependency org="commons-logging"
name="commons-logging"
rev="${commons-logging.version}"
Modified:
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacitySchedulerConf.java
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacitySchedulerConf.java?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
---
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacitySchedulerConf.java
(original)
+++
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacitySchedulerConf.java
Tue Jun 9 16:11:19 2009
@@ -164,10 +164,10 @@
* Sets the capacity of the given queue.
*
* @param queue name of the queue
- * @param gc percent of the cluster for the queue.
+ * @param capacity percent of the cluster for the queue.
*/
- public void setCapacity(String queue,float gc) {
- rmConf.setFloat(toFullPropertyName(queue, "capacity"),gc);
+ public void setCapacity(String queue,float capacity) {
+ rmConf.setFloat(toFullPropertyName(queue, "capacity"),capacity);
}
/**
Modified:
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacityTaskScheduler.java
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacityTaskScheduler.java?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
---
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacityTaskScheduler.java
(original)
+++
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/CapacityTaskScheduler.java
Tue Jun 9 16:11:19 2009
@@ -24,8 +24,6 @@
import java.util.Comparator;
import java.util.HashMap;
import java.util.HashSet;
-import java.util.Iterator;
-import java.util.LinkedList;
import java.util.List;
import java.util.Map;
import java.util.Set;
@@ -34,8 +32,6 @@
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.mapred.JobTracker.IllegalStateException;
-import org.apache.hadoop.util.StringUtils;
-
/**
* A {...@link TaskScheduler} that implements the requirements in HADOOP-3421
@@ -78,14 +74,20 @@
private static class TaskSchedulingInfo {
/**
- * the actual gc, which depends on how many slots are available
+ * the actual capacity, which depends on how many slots are available
* in the cluster at any given time.
*/
int capacity = 0;
// number of running tasks
int numRunningTasks = 0;
- /** for each user, we need to keep track of number of running tasks */
- Map<String, Integer> numRunningTasksByUser =
+ // number of slots occupied by running tasks
+ int numSlotsOccupied = 0;
+
+ /**
+ * for each user, we need to keep track of number of slots occupied by
+ * running tasks
+ */
+ Map<String, Integer> numSlotsOccupiedByUser =
new HashMap<String, Integer>();
/**
@@ -93,32 +95,41 @@
*/
void resetTaskVars() {
numRunningTasks = 0;
- for (String s: numRunningTasksByUser.keySet()) {
- numRunningTasksByUser.put(s, 0);
+ numSlotsOccupied = 0;
+ for (String s: numSlotsOccupiedByUser.keySet()) {
+ numSlotsOccupiedByUser.put(s, Integer.valueOf(0));
}
}
/**
* return information about the tasks
*/
- public String toString(){
- float runningTasksAsPercent = capacity!= 0 ?
- ((float)numRunningTasks * 100/capacity):0;
+ @Override
+ public String toString() {
+ float occupiedSlotsAsPercent =
+ capacity != 0 ? ((float) numSlotsOccupied * 100 / capacity) : 0;
StringBuffer sb = new StringBuffer();
- sb.append("Capacity: " + capacity + "\n");
- sb.append(String.format("Running tasks: %.1f%% of Capacity\n",
- runningTasksAsPercent));
+ sb.append("Capacity: " + capacity + " slots\n");
+ sb.append(String.format("Used capacity: %d (%.1f%% of Capacity)\n",
+ Integer.valueOf(numSlotsOccupied), Float
+ .valueOf(occupiedSlotsAsPercent)));
+ sb.append(String.format("Running tasks: %d\n", Integer
+ .valueOf(numRunningTasks)));
// include info on active users
- if (numRunningTasks != 0) {
+ if (numSlotsOccupied != 0) {
sb.append("Active users:\n");
- for (Map.Entry<String, Integer> entry:
numRunningTasksByUser.entrySet()) {
+ for (Map.Entry<String, Integer> entry : numSlotsOccupiedByUser
+ .entrySet()) {
if ((entry.getValue() == null) || (entry.getValue().intValue() <=
0)) {
// user has no tasks running
continue;
}
- sb.append("User '" + entry.getKey()+ "': ");
- float p = (float)entry.getValue().intValue()*100/numRunningTasks;
- sb.append(String.format("%.1f%% of running tasks\n", p));
+ sb.append("User '" + entry.getKey() + "': ");
+ int numSlotsOccupiedByThisUser = entry.getValue().intValue();
+ float p =
+ (float) numSlotsOccupiedByThisUser * 100 / numSlotsOccupied;
+ sb.append(String.format("%d (%.1f%% of used capacity)\n", Long
+ .valueOf(numSlotsOccupiedByThisUser), Float.valueOf(p)));
}
}
return sb.toString();
@@ -152,10 +163,10 @@
TaskSchedulingInfo mapTSI;
TaskSchedulingInfo reduceTSI;
- public QueueSchedulingInfo(String queueName, float gcPercent,
+ public QueueSchedulingInfo(String queueName, float capacityPercent,
int ulMin, JobQueuesManager jobQueuesManager) {
this.queueName = new String(queueName);
- this.capacityPercent = gcPercent;
+ this.capacityPercent = capacityPercent;
this.ulMin = ulMin;
this.jobQueuesManager = jobQueuesManager;
this.mapTSI = new TaskSchedulingInfo();
@@ -164,13 +175,14 @@
/**
* return information about the queue
+ * @return a String representing the information about the queue.
*/
+ @Override
public String toString(){
// We print out the queue information first, followed by info
// on map and reduce tasks and job info
StringBuffer sb = new StringBuffer();
sb.append("Queue configuration\n");
- //sb.append("Name: " + queueName + "\n");
sb.append("Capacity Percentage: ");
sb.append(capacityPercent);
sb.append("%\n");
@@ -281,7 +293,15 @@
protected CapacityTaskScheduler.TYPE type = null;
abstract Task obtainNewTask(TaskTrackerStatus taskTracker,
- JobInProgress job) throws IOException;
+ JobInProgress job) throws IOException;
+
+ int getSlotsOccupied(JobInProgress job) {
+ return getRunningTasks(job) * getSlotsPerTask(job);
+ }
+
+ abstract int getClusterCapacity();
+ abstract int getSlotsPerTask(JobInProgress job);
+ abstract int getRunningTasks(JobInProgress job);
abstract int getPendingTasks(JobInProgress job);
abstract TaskSchedulingInfo getTSI(QueueSchedulingInfo qsi);
/**
@@ -313,12 +333,12 @@
public int compare(QueueSchedulingInfo q1, QueueSchedulingInfo q2) {
TaskSchedulingInfo t1 = getTSI(q1);
TaskSchedulingInfo t2 = getTSI(q2);
- // look at how much capacity they've filled. Treat a queue with gc=0
- // equivalent to a queue running at capacity
+ // look at how much capacity they've filled. Treat a queue with
+ // capacity=0 equivalent to a queue running at capacity
double r1 = (0 == t1.capacity)? 1.0f:
- (double)t1.numRunningTasks/(double)t1.capacity;
+ (double)t1.numSlotsOccupied/(double)t1.capacity;
double r2 = (0 == t2.capacity)? 1.0f:
- (double)t2.numRunningTasks/(double)t2.capacity;
+ (double)t2.numSlotsOccupied/(double)t2.capacity;
if (r1<r2) return -1;
else if (r1>r2) return 1;
else return 0;
@@ -340,7 +360,17 @@
protected final static ReduceQueueComparator reduceComparator = new
ReduceQueueComparator();
// and this is the comparator to use
protected QueueComparator queueComparator;
-
+
+ // Returns queues sorted according to the QueueComparator.
+ // Mainly for testing purposes.
+ String[] getOrderedQueues() {
+ List<String> queues = new ArrayList<String>(qsiForAssigningTasks.size());
+ for (QueueSchedulingInfo qsi : qsiForAssigningTasks) {
+ queues.add(qsi.queueName);
+ }
+ return queues.toArray(new String[queues.size()]);
+ }
+
TaskSchedulingMgr(CapacityTaskScheduler sched) {
scheduler = sched;
}
@@ -357,24 +387,26 @@
}
- private boolean isUserOverLimit(String user, QueueSchedulingInfo qsi) {
- // what is our current capacity? It's capacity if we're running below
capacity.
- // If we're running over capacity, then its #running plus 1 (which is the
- // extra slot we're getting).
+ private boolean isUserOverLimit(JobInProgress j, QueueSchedulingInfo qsi) {
+ // what is our current capacity? It is equal to the queue-capacity if
+ // we're running below capacity. If we're running over capacity, then its
+ // #running plus slotPerTask of the job (which is the number of extra
+ // slots we're getting).
int currentCapacity;
TaskSchedulingInfo tsi = getTSI(qsi);
- if (tsi.numRunningTasks < tsi.capacity) {
+ if (tsi.numSlotsOccupied < tsi.capacity) {
currentCapacity = tsi.capacity;
}
else {
- currentCapacity = tsi.numRunningTasks+1;
+ currentCapacity = tsi.numSlotsOccupied + getSlotsPerTask(j);
}
int limit = Math.max((int)(Math.ceil((double)currentCapacity/
(double)qsi.numJobsByUser.size())),
(int)(Math.ceil((double)(qsi.ulMin*currentCapacity)/100.0)));
- if (tsi.numRunningTasksByUser.get(user) >= limit) {
- LOG.debug("User " + user + " is over limit, num running tasks = " +
- tsi.numRunningTasksByUser.get(user) + ", limit = " + limit);
+ String user = j.getProfile().getUser();
+ if (tsi.numSlotsOccupiedByUser.get(user) >= limit) {
+ LOG.debug("User " + user + " is over limit, num slots occupied = " +
+ tsi.numSlotsOccupiedByUser.get(user) + ", limit = " + limit);
return true;
}
else {
@@ -403,7 +435,7 @@
continue;
}
// check if the job's user is over limit
- if (isUserOverLimit(j.getProfile().getUser(), qsi)) {
+ if (isUserOverLimit(j, qsi)) {
continue;
}
//If this job meets memory requirements. Ask the JobInProgress for
@@ -489,8 +521,11 @@
// The caller is responsible for ensuring that the QSI objects and the
// collections are up-to-date.
private TaskLookupResult assignTasks(TaskTrackerStatus taskTracker) throws
IOException {
+
+ printQSIs();
+
for (QueueSchedulingInfo qsi : qsiForAssigningTasks) {
- // we may have queues with gc=0. We shouldn't look at jobs from
+ // we may have queues with capacity=0. We shouldn't look at jobs from
// these queues
if (0 == getTSI(qsi).capacity) {
continue;
@@ -516,20 +551,23 @@
// nothing to give
return TaskLookupResult.getNoTaskFoundResult();
}
-
+
// for debugging.
private void printQSIs() {
- StringBuffer s = new StringBuffer();
- for (QueueSchedulingInfo qsi: qsiForAssigningTasks) {
- TaskSchedulingInfo tsi = getTSI(qsi);
- Collection<JobInProgress> runJobs =
- scheduler.jobQueuesManager.getRunningJobQueue(qsi.queueName);
- s.append(" Queue '" + qsi.queueName + "'(" + this.type + "): run=" +
- tsi.numRunningTasks + ", gc=" + tsi.capacity
- + ", run jobs="+ runJobs.size() +
- "*** ");
+ if (LOG.isDebugEnabled()) {
+ StringBuffer s = new StringBuffer();
+ for (QueueSchedulingInfo qsi : qsiForAssigningTasks) {
+ TaskSchedulingInfo tsi = getTSI(qsi);
+ Collection<JobInProgress> runJobs =
+ scheduler.jobQueuesManager.getRunningJobQueue(qsi.queueName);
+ s.append(String.format(" Queue '%s'(%s): runningTasks=%d, "
+ + "occupiedSlots=%d, capacity=%d, runJobs=%d", qsi.queueName,
+ this.type, Integer.valueOf(tsi.numRunningTasks), Integer
+ .valueOf(tsi.numSlotsOccupied), Integer
+ .valueOf(tsi.capacity), Integer.valueOf(runJobs.size())));
+ }
+ LOG.debug(s);
}
- LOG.debug(s);
}
/**
@@ -559,11 +597,14 @@
* The scheduling algorithms for map tasks.
*/
private static class MapSchedulingMgr extends TaskSchedulingMgr {
- MapSchedulingMgr(CapacityTaskScheduler dad) {
- super(dad);
+
+ MapSchedulingMgr(CapacityTaskScheduler schedulr) {
+ super(schedulr);
type = CapacityTaskScheduler.TYPE.MAP;
queueComparator = mapComparator;
}
+
+ @Override
Task obtainNewTask(TaskTrackerStatus taskTracker, JobInProgress job)
throws IOException {
ClusterStatus clusterStatus =
@@ -572,16 +613,30 @@
return job.obtainNewMapTask(taskTracker, numTaskTrackers,
scheduler.taskTrackerManager.getNumberOfUniqueHosts());
}
+
+ @Override
int getClusterCapacity() {
return scheduler.taskTrackerManager.getClusterStatus().getMaxMapTasks();
}
+
+ @Override
int getRunningTasks(JobInProgress job) {
return job.runningMaps();
}
+
+ @Override
int getPendingTasks(JobInProgress job) {
return job.pendingMaps();
}
+ @Override
+ int getSlotsPerTask(JobInProgress job) {
+ long myVmem = job.getJobConf().getMemoryForMapTask();
+ return (int) (Math.ceil((float) myVmem
+ / (float) scheduler.getMemSizeForMapSlot()));
+ }
+
+ @Override
TaskSchedulingInfo getTSI(QueueSchedulingInfo qsi) {
return qsi.mapTSI;
}
@@ -601,11 +656,14 @@
* The scheduling algorithms for reduce tasks.
*/
private static class ReduceSchedulingMgr extends TaskSchedulingMgr {
- ReduceSchedulingMgr(CapacityTaskScheduler dad) {
- super(dad);
+
+ ReduceSchedulingMgr(CapacityTaskScheduler schedulr) {
+ super(schedulr);
type = CapacityTaskScheduler.TYPE.REDUCE;
queueComparator = reduceComparator;
}
+
+ @Override
Task obtainNewTask(TaskTrackerStatus taskTracker, JobInProgress job)
throws IOException {
ClusterStatus clusterStatus =
@@ -614,16 +672,31 @@
return job.obtainNewReduceTask(taskTracker, numTaskTrackers,
scheduler.taskTrackerManager.getNumberOfUniqueHosts());
}
+
+ @Override
int getClusterCapacity() {
- return
scheduler.taskTrackerManager.getClusterStatus().getMaxReduceTasks();
+ return scheduler.taskTrackerManager.getClusterStatus()
+ .getMaxReduceTasks();
}
+
+ @Override
int getRunningTasks(JobInProgress job) {
return job.runningReduces();
}
+
+ @Override
int getPendingTasks(JobInProgress job) {
return job.pendingReduces();
}
+ @Override
+ int getSlotsPerTask(JobInProgress job) {
+ long myVmem = job.getJobConf().getMemoryForReduceTask();
+ return (int) (Math.ceil((float) myVmem
+ / (float) scheduler.getMemSizeForReduceSlot()));
+ }
+
+ @Override
TaskSchedulingInfo getTSI(QueueSchedulingInfo qsi) {
return qsi.reduceTSI;
}
@@ -655,7 +728,10 @@
protected CapacitySchedulerConf schedConf;
/** whether scheduler has started or not */
private boolean started = false;
-
+
+ static String JOB_SCHEDULING_INFO_FORMAT_STRING =
+ "%s running map tasks using %d map slots,"
+ + " %s running reduce tasks using %d reduce slots.";
/**
* A clock class - can be mocked out for testing.
*/
@@ -664,6 +740,7 @@
return System.currentTimeMillis();
}
}
+
// can be replaced with a global type, if we have one
protected static enum TYPE {
MAP, REDUCE
@@ -709,12 +786,13 @@
JobConf.normalizeMemoryConfigValue(conf.getLong(
JobTracker.MAPRED_CLUSTER_MAX_REDUCE_MEMORY_MB_PROPERTY,
JobConf.DISABLED_MEMORY_LIMIT));
- LOG.info(new StringBuilder().append("Scheduler configured with ").append(
- "(memSizeForMapSlotOnJT, memSizeForReduceSlotOnJT,").append(
- " limitMaxMemForMapTasks, limitMaxMemForReduceTasks)").append(
- memSizeForMapSlotOnJT).append(", ").append(memSizeForReduceSlotOnJT)
- .append(", ").append(limitMaxMemForMapTasks).append(", ").append(
- limitMaxMemForReduceTasks).append(")"));
+ LOG.info(String.format("Scheduler configured with "
+ + "(memSizeForMapSlotOnJT, memSizeForReduceSlotOnJT, "
+ + "limitMaxMemForMapTasks, limitMaxMemForReduceTasks)"
+ + " (%d,%d,%d,%d)", Long.valueOf(memSizeForMapSlotOnJT), Long
+ .valueOf(memSizeForReduceSlotOnJT), Long
+ .valueOf(limitMaxMemForMapTasks), Long
+ .valueOf(limitMaxMemForReduceTasks)));
}
long getMemSizeForMapSlot() {
@@ -733,6 +811,15 @@
return limitMaxMemForReduceTasks;
}
+ String[] getOrderedQueues(CapacityTaskScheduler.TYPE type) {
+ if (type.equals(CapacityTaskScheduler.TYPE.MAP)) {
+ return mapScheduler.getOrderedQueues();
+ } else if (type.equals(CapacityTaskScheduler.TYPE.REDUCE)) {
+ return reduceScheduler.getOrderedQueues();
+ }
+ return null;
+ }
+
@Override
public synchronized void start() throws IOException {
if (started) return;
@@ -755,15 +842,15 @@
Set<String> queuesWithoutConfiguredCapacity = new HashSet<String>();
float totalCapacity = 0.0f;
for (String queueName: queues) {
- float gc = schedConf.getCapacity(queueName);
- if(gc == -1.0) {
+ float capacity = schedConf.getCapacity(queueName);
+ if(capacity == -1.0) {
queuesWithoutConfiguredCapacity.add(queueName);
}else {
- totalCapacity += gc;
+ totalCapacity += capacity;
}
int ulMin = schedConf.getMinimumUserLimitPercent(queueName);
// create our QSI and add to our hashmap
- QueueSchedulingInfo qsi = new QueueSchedulingInfo(queueName, gc,
+ QueueSchedulingInfo qsi = new QueueSchedulingInfo(queueName, capacity,
ulMin, jobQueuesManager);
queueInfoMap.put(queueName, qsi);
@@ -877,29 +964,49 @@
if (j.getStatus().getRunState() != JobStatus.RUNNING) {
continue;
}
- int runningMaps = j.runningMaps();
- int runningReduces = j.runningReduces();
- qsi.mapTSI.numRunningTasks += runningMaps;
- qsi.reduceTSI.numRunningTasks += runningReduces;
+
+ int numMapsRunningForThisJob = mapScheduler.getRunningTasks(j);
+ int numReducesRunningForThisJob = reduceScheduler.getRunningTasks(j);
+ int numMapSlotsForThisJob = mapScheduler.getSlotsOccupied(j);
+ int numReduceSlotsForThisJob = reduceScheduler.getSlotsOccupied(j);
+ j.setSchedulingInfo(String.format(JOB_SCHEDULING_INFO_FORMAT_STRING,
+ Integer.valueOf(numMapsRunningForThisJob), Integer
+ .valueOf(numMapSlotsForThisJob), Integer
+ .valueOf(numReducesRunningForThisJob), Integer
+ .valueOf(numReduceSlotsForThisJob)));
+ qsi.mapTSI.numRunningTasks += numMapsRunningForThisJob;
+ qsi.reduceTSI.numRunningTasks += numReducesRunningForThisJob;
+ qsi.mapTSI.numSlotsOccupied += numMapSlotsForThisJob;
+ qsi.reduceTSI.numSlotsOccupied += numReduceSlotsForThisJob;
Integer i =
- qsi.mapTSI.numRunningTasksByUser.get(j.getProfile().getUser());
- qsi.mapTSI.numRunningTasksByUser.put(j.getProfile().getUser(),
- i+runningMaps);
- i = qsi.reduceTSI.numRunningTasksByUser.get(j.getProfile().getUser());
- qsi.reduceTSI.numRunningTasksByUser.put(j.getProfile().getUser(),
- i+runningReduces);
- LOG.debug("updateQSI: job " + j.getJobID().toString() + ": run(m) = " +
- j.runningMaps() + ", run(r) = " + j.runningReduces() +
- ", finished(m) = " + j.finishedMaps() + ", finished(r)= " +
- j.finishedReduces() + ", failed(m) = " + j.failedMapTasks +
- ", failed(r) = " + j.failedReduceTasks + ", spec(m) = " +
- j.speculativeMapTasks + ", spec(r) = " + j.speculativeReduceTasks
- + ", total(m) = " + j.numMapTasks + ", total(r) = " +
- j.numReduceTasks);
+ qsi.mapTSI.numSlotsOccupiedByUser.get(j.getProfile().getUser());
+ qsi.mapTSI.numSlotsOccupiedByUser.put(j.getProfile().getUser(),
+ Integer.valueOf(i.intValue() + numMapSlotsForThisJob));
+ i = qsi.reduceTSI.numSlotsOccupiedByUser.get(j.getProfile().getUser());
+ qsi.reduceTSI.numSlotsOccupiedByUser.put(j.getProfile().getUser(),
+ Integer.valueOf(i.intValue() + numReduceSlotsForThisJob));
+ if (LOG.isDebugEnabled()) {
+ LOG.debug(String.format("updateQSI: job %s: run(m)=%d, "
+ + "occupied(m)=%d, run(r)=%d, occupied(r)=%d, finished(m)=%d,"
+ + " finished(r)=%d, failed(m)=%d, failed(r)=%d, "
+ + "spec(m)=%d, spec(r)=%d, total(m)=%d, total(r)=%d", j
+ .getJobID().toString(), Integer
+ .valueOf(numMapsRunningForThisJob), Integer
+ .valueOf(numMapSlotsForThisJob), Integer
+ .valueOf(numReducesRunningForThisJob), Integer
+ .valueOf(numReduceSlotsForThisJob), Integer.valueOf(j
+ .finishedMaps()), Integer.valueOf(j.finishedReduces()), Integer
+ .valueOf(j.failedMapTasks),
+ Integer.valueOf(j.failedReduceTasks), Integer
+ .valueOf(j.speculativeMapTasks), Integer
+ .valueOf(j.speculativeReduceTasks), Integer
+ .valueOf(j.numMapTasks), Integer.valueOf(j.numReduceTasks)));
+ }
+
/*
* it's fine walking down the entire list of running jobs - there
* probably will not be many, plus, we may need to go through the
- * list to compute numRunningTasksByUser. If this is expensive, we
+ * list to compute numSlotsOccupiedByUser. If this is expensive, we
* can keep a list of running jobs per user. Then we only need to
* consider the first few jobs per user.
*/
@@ -914,10 +1021,8 @@
* The grand plan for assigning a task.
* First, decide whether a Map or Reduce task should be given to a TT
* (if the TT can accept either).
- * Next, pick a queue. We only look at queues that need a slot. Among
- * these, we first look at queues whose ac is less than gc (queues that
- * gave up capacity in the past). Next, we look at any other queue that
- * needs a slot.
+ * Next, pick a queue. We only look at queues that need a slot. Among these,
+ * we first look at queues whose (# of running tasks)/capacity is the least.
* Next, pick a job in a queue. we pick the job at the front of the queue
* unless its user is over the user limit.
* Finally, given a job, pick a task from the job.
@@ -1019,8 +1124,10 @@
if (null == i) {
i = 1;
// set the count for running tasks to 0
- qsi.mapTSI.numRunningTasksByUser.put(job.getProfile().getUser(), 0);
- qsi.reduceTSI.numRunningTasksByUser.put(job.getProfile().getUser(), 0);
+ qsi.mapTSI.numSlotsOccupiedByUser.put(job.getProfile().getUser(),
+ Integer.valueOf(0));
+ qsi.reduceTSI.numSlotsOccupiedByUser.put(job.getProfile().getUser(),
+ Integer.valueOf(0));
}
else {
i++;
@@ -1042,8 +1149,8 @@
if (0 == i.intValue()) {
qsi.numJobsByUser.remove(job.getProfile().getUser());
// remove job footprint from our TSIs
- qsi.mapTSI.numRunningTasksByUser.remove(job.getProfile().getUser());
- qsi.reduceTSI.numRunningTasksByUser.remove(job.getProfile().getUser());
+ qsi.mapTSI.numSlotsOccupiedByUser.remove(job.getProfile().getUser());
+ qsi.reduceTSI.numSlotsOccupiedByUser.remove(job.getProfile().getUser());
LOG.debug("No more jobs for user, number of users = " +
qsi.numJobsByUser.size());
}
else {
Modified:
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/MemoryMatcher.java
URL:
http://svn.apache.org/viewvc/hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/MemoryMatcher.java?rev=783055&r1=783054&r2=783055&view=diff
==============================================================================
---
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/MemoryMatcher.java
(original)
+++
hadoop/core/branches/HADOOP-3628-2/src/contrib/capacity-scheduler/src/java/org/apache/hadoop/mapred/MemoryMatcher.java
Tue Jun 9 16:11:19 2009
@@ -53,10 +53,9 @@
* @return amount of memory that is used by the residing tasks,
* null if memory cannot be computed for some reason.
*/
- private synchronized Long getMemReservedForTasks(
+ synchronized Long getMemReservedForTasks(
TaskTrackerStatus taskTracker, CapacityTaskScheduler.TYPE taskType) {
long vmem = 0;
- long myVmem = 0;
for (TaskStatus task : taskTracker.getTaskReports()) {
// the following task states are one in which the slot is
@@ -90,6 +89,7 @@
// tasks' memory limits to the nearest multiple of the slot-memory-size
// set on JT. This essentially translates to tasks of a high memory job
// using multiple slots.
+ long myVmem = 0;
if (task.getIsMap() &&
taskType.equals(CapacityTaskScheduler.TYPE.MAP)) {
myVmem = jConf.getMemoryForMapTask();
myVmem =
