Below is a compilation of all changes that were made in the Cassandra code base related to Hadoop support. The source for this compilation is Apache Cassandra - CHANGES.txt. I have tried my best to avoid any misses or mistakes. In case you notice something amiss, please drop in a comment and I will fix it.

1.0.1

  • Skip empty rows when slicing the entire row (CASSANDRA-2855)
  • Make CFIF try rpc_address or fallback to listen_address (CASSANDRA-3214)
  • Accept comma delimited lists of initial thrift connections (CASSANDRA-3158)

0.8.7

0.8.5

  • Fail jobs when Cassandra node has failed but TaskTracker has not (CASSANDRA-2388)

0.8.3

0.8.2

0.8.1

  • Fix race that could result in Hadoop writer failing to throw an exception encountered after close (CASSANDRA-2755)

0.8.0

0.7.5

  • Allow job configuration to set the CL used in Hadoop jobs (CASSANDRA-2331)

0.7.3

  • Fix Hadoop ColumnFamilyOutputFormat dropping of mutations when batch fills up (CASSANDRA-2255)

0.7.1

0.7.0-rc2

  • Support multiple Mutations per key in hadoop ColumnFamilyOutputFormat (CASSANDRA-1774)

0.7-beta2

  • Remove cassandra.yaml dependency from Hadoop and Pig (CASSANDRA-1322)
  • Support for Hadoop Streaming [non-jvm map/reduce via stdin/out] (CASSANDRA-1368)
  • Rewrite Hadoop ColumnFamilyRecordWriter to pool connections, retry to multiple Cassandra nodes, and smooth impact on the Cassandra cluster by using smaller batch sizes (CASSANDRA-1434)

0.7-beta1

0.6.4

  • Hadoop jobs no longer require the Cassandra storage-conf.xml (CASSANDRA-1047)

0.6.2

  • Fix SlicePredicate serialization inside Hadoop jobs (CASSANDRA-1049)
  • Close Thrift sockets in Hadoop ColumnFamilyRecordReader (CASSANDRA-1081)

0.6.1

  • Use hostnames in CFInputFormat to allow Hadoop’s naive string-based locality comparisons to work (CASSANDRA-955)

0.6.0-beta3

0.6.0-beta1/beta2