This is the documentation for CDH 4.7.1.
Documentation for other versions is available at Cloudera Documentation.

What's New in CDH4.2.1

This is a maintenance release that fixes a number of issues including the following.

Apache HDFS

  • HDFS-4618 - Default transaction interval for checkpoints is too low
  • HADOOP-9125 - LdapGroupsMapping threw CommunicationException after some idle time.
  • HADOOP-9150 - Avoid unnecessary DNS resolution attempts for logical URIs
  • HDFS-4595 - When short circuit read is fails, DFSClient does not fallback to regular read
  • HDFS-4538 - Allow use of legacy blockreader
  • HDFS-4569 - Small image transfer related cleanups
  • HDFS-3224 - Bug in check for DN re-registration with different storage ID
  • HDFS-3990 - NN's health report has severe performance problems
  • HADOOP-8469 - Make NetworkTopology class pluggable
  • HDFS-4246 - The exclude node list should be more forgiving, for each output stream
  • HDFS-4521 - Invalid network toploogies should not be cached
  • HDFS-4304 - Make FSEditLogOp.MAX_OP_SIZE configurable
  • HDFS-3277 - Fail over to loading a different FSImage if the first one we try to load is corrupt.
  • HADOOP-5442 - The job history display needs to be paged
  • HADOOP-7108 - JobConf link broken in job history page
  • HDFS-4596 - Shutting down namenode during checkpointing can lead to md5sum error
  • HDFS-4591 - HA clients can fail to fail over while Standby NN is performing long checkpoint
  • HADOOP-8159 - Mismatch of topology script mappings can trigger HADOOP-8159 if hostname/IP mapping mismatches for the same DN

Apache MapReduce

  • MAPREDUCE-5008 - Merger progress miscounts with respect to EOF_MARKER
  • MAPREDUCE-2217 - (Fixed case in which this bug leaves orphan processes hanging)
  • MAPREDUCE-4843 - When using DefaultTaskController, JobLocalizer not thread safe
  • HADOOP-5442 - The job history display needs to be paged
  • MAPREDUCE-4888 - NLineInputFormat drops data in 1.1 and beyond

Apache HBase

  • HBASE-8199 - Eliminate exception for ExportSnapshot against the null table snapshot
  • HBASE-8099 - ReplicationZookeeper.copyQueuesFromRSUsingMulti should not return any queues if it failed to execute.
  • HBASE-7991 - HFileReaderV1 caching the same parent META block could cause server abort when splitting
  • HBASE-8211 - Support for NN HA for 0.94
  • HBASE-7369 - HConnectionManager should remove aborted connections
  • HBASE-8211 - Support for NN HA for 0.94
  • HBASE-8288 - HBaseFileSystem: Refactoring and correct semantics for createPath methods

Apache Hive

  • HIVE-4141 - InspectorFactories contains static HashMaps which can cause infinite loop
  • HIVE-4075 - TypeInfoFactory is not thread safe and is accessed by multiple threads
  • HIVE-2264 - Hive Driver calls System.exit()which causes HiveServer to exit
  • HIVE-4122 - Queries fail if timestamp data not in expected format
  • HIVE-3528 - Avro SerDe doesn't handle serializing Nullable types that require access to a schema
  • HIVE-4119 - ANALYZE TABLE ... COMPUTE STATISTICS FOR COLUMNS fails with NPE if the table is empty
  • HIVE-4001 - Add o.a.h.h.serde.Constants for backward compatibility

Hue

  • Support for nested groups for LDAP import in Hue
  • HUE-1082 - LDAP Sync Users fails in the latest code branch
  • HUE-1080 - LDAP Sync Users fails with AD in Hue 2.2
  • HUE-1064 - Hue doesn't work when hadoop.rpc.protection is set to anything other than authentication