Apache Hadoop Changelog

Release 0.23.4 - 2012-10-15

NEW FEATURES:

JIRA Summary Priority Component Reporter Contributor
HADOOP-7788 HA: Simple HealthMonitor class to watch an HAService Major ha Todd Lipcon Todd Lipcon
MAPREDUCE-4651 Benchmarking random reads with DFSIO Major benchmarks, test Konstantin Shvachko Konstantin Shvachko

IMPROVEMENTS:

JIRA Summary Priority Component Reporter Contributor
HADOOP-8183 Stop using “mapred.used.genericoptionsparser” to avoid unnecessary warnings Minor util Harsh J Harsh J
HADOOP-8623 hadoop jar command should respect HADOOP_OPTS Minor scripts Steven Willis Steven Willis
MAPREDUCE-4408 allow jobs to set a JAR that is in the distributed cached Major mrv1, mrv2 Alejandro Abdelnur Robert Kanter
MAPREDUCE-2786 TestDFSIO should also test compression reading/writing from command-line. Minor benchmarks Plamen Jeliazkov Plamen Jeliazkov
YARN-57 Plugable process tree Major nodemanager Radim Kolar Radim Kolar
MAPREDUCE-4645 Providing a random seed to Slive should make the sequence of filenames completely deterministic Major performance, test Ravi Prakash Ravi Prakash
YARN-137 Change the default scheduler to the CapacityScheduler Major scheduler Siddharth Seth Siddharth Seth

BUG FIXES:

JIRA Summary Priority Component Reporter Contributor
MAPREDUCE-4193 broken doc link for yarn-default.xml in site.xml Major documentation Patrick Hunt Patrick Hunt
HADOOP-8310 FileContext#checkPath should handle URIs with no port Major fs Aaron T. Myers Aaron T. Myers
HDFS-3860 HeartbeatManager#Monitor may wrongly hold the writelock of namesystem Major . Jing Zhao Jing Zhao
HADOOP-8684 Deadlock between WritableComparator and WritableComparable Minor io Hiroshi Ikeda Jing Zhao
YARN-42 Node Manager throws NPE on startup Major nodemanager Devaraj K Devaraj K
YARN-75 RMContainer should handle a RELEASE event while RUNNING Major . Siddharth Seth Siddharth Seth
MAPREDUCE-4646 client does not receive job diagnostics for failed jobs Major mrv2 Jason Lowe Jason Lowe
YARN-93 Diagnostics missing from applications that have finished but failed Major resourcemanager Jason Lowe Jason Lowe
YARN-108 FSDownload can create cache directories with the wrong permissions Critical nodemanager Jason Lowe Jason Lowe
YARN-88 DefaultContainerExecutor can fail to set proper permissions Major nodemanager Jason Lowe Jason Lowe
HADOOP-8822 relnotes.py was deleted post mavenization Major . Robert Joseph Evans Robert Joseph Evans
MAPREDUCE-4647 We should only unjar jobjar if there is a lib directory in it. Major mrv2 Robert Joseph Evans Robert Joseph Evans
HADOOP-8843 Old trash directories are never deleted on upgrade from 1.x Critical . Robert Joseph Evans Jason Lowe
HDFS-3831 Failure to renew tokens due to test-sources left in classpath Critical security Jason Lowe Jason Lowe
HDFS-3922 0.22 and 0.23 namenode throws away blocks under construction on restart Critical namenode Kihwal Lee Kihwal Lee
HDFS-3731 2.0 release upgrade must handle blocks being written from 1.0 Blocker datanode Suresh Srinivas Kihwal Lee
HDFS-3373 FileContext HDFS implementation can leak socket caches Major hdfs-client Todd Lipcon John George
YARN-106 Nodemanager needs to set permissions of local directories Major nodemanager Jason Lowe Jason Lowe
MAPREDUCE-4691 Historyserver can report “Unknown job” after RM says job has completed Critical jobhistoryserver, mrv2 Jason Lowe Robert Joseph Evans
MAPREDUCE-4689 JobClient.getMapTaskReports on failed job results in NPE Major client Jason Lowe Jason Lowe
YARN-138 Improve default config values for YARN Major resourcemanager, scheduler Arun C Murthy Harsh J
HDFS-3626 Creating file with invalid path can corrupt edit log Blocker namenode Todd Lipcon Todd Lipcon
HDFS-3553 Hftp proxy tokens are broken Blocker . Daryn Sharp Daryn Sharp