DataStax Enterprise 2.1 Documentation

Release Notes

DataStax Enterprise 2.1.1

This release fixes two issues that cause an exception that looks something like this:

java.lang.ArrayIndexOutOfBoundsException: 0
   at com.datastax.bdp.hadoop.cfs.CassandraFileSystem.mkdir . . .

  • Issue 1: Array Index Out-Of-Bounds if Hadoop UserGroupInformation (UGI) returns no groups for the user

    Resolves the problem caused by directory and file creation when running Mahout.

  • Issue 2: Array Index Out-Of-Bounds if checking permissions even when dfs.permissions is set to false on all nodes

    Resolves the problem of the dfs.permissions setting in core-site.xml being ignored under certain conditions. For example, when submitting a remote job from a user that does not exist on an Analytics node, dfs.permissions set to false did not turn off permission checking as expected. User/group information code continued to be returned, resulting in an exception.

DataStax Enterprise 2.1.0

  • Apache Cassandra 1.0.10
  • Apache Hadoop 1.0.0
  • Apache Hive 0.8.1
  • Apache Mahout 0.6
  • Apache Pig 0.8.3
  • Apache Sqoop 1.4.1
  • Apache Solr 4.0
  • DataStax OpsCenter 2.1

What's New

DataStax Enterprise 2.1 extended the DataStax big data platform to include:

  • Enhanced multi-data center Hadoop support, - includes the capability to run multiple job trackers across multiple data centers and create multiple Cassandra File System (CFS) keyspaces per data center, which keeps metadata local to each data center for faster performance.
  • Mahout support - includes machine learning and data mining capabilities in their analytics applications.
  • Oracle Unbreakable Linux support.
  • Support for Multiple CassandraFS's - In DataStax Enterprise 2.1 and later, you can create additional CassandraFS's to organize Hadoop data. For example, using multiple CassandraFS's you can set different replication options for keyspaces.

DataStax Enterprise 2.0

  • Apache Cassandra 1.0.8
  • Apache Hadoop 1.0.0
  • Apache Hive 0.8.1
  • Apache Pig 0.8.3
  • Apache Sqoop 1.4.1
  • Apache Solr 4.0
  • DataStax OpsCenter 2.0

What's New

With version 2.0 of DataStax Enterprise, DataStax has extended its big data platform to include:

  • Integration with Apache Solr™ - Enables DataStax Enterprise to use full-text search; hit highlighting; multiple search attributes; search rich documents, such as PDF and Microsoft Word; and use geo-spatial search. By including Solr 4.0, near real-time indexing can be performed to manage real-time, analytics, and enterprise search features within a single integrated platform.
  • Hadoop Update - DataStax Enterprise 2.0 includes Hadoop 1.0.0.
  • Improved Hive Support - DataStax Enterprise is now enabled for any JDBC compliant user interface to connect to and work with Hive from within the server. This version of Hive includes support for binary data and support for wide rows (up to 2 billion columns).
  • Sqoop - Apache Sqoop supports migration of RDBMS data into the DataStax Enterprise server. You can now easily import data from RDBMS’s such as Oracle, Microsoft SQL Server, MySQL, Sybase, DB2, and others.
  • Elastic Workload Re-provisioning - Provides the ability to re-provision existing nodes to assume a different workload, such as changing a real-time node to an analytics node, and the ability to change the overall usage and capacity of a cluster.
  • Log4j - Apache log4j is a Java-based logging framework that provides runtime application feedback. It provides the ability to control the granularity of log statements using an external configuration file. Additionally, with the Cassandra Appender you can store the log4j messages in a column family where they're available for in-depth analysis using the Hadoop and Solr capabilities provided by DataStax Enterprise.

For more detailed information, see About DataStax Enterprise and What’s New in DataStax Enterprise 2.0? white paper.

Other issues

Do not use SolrJ (Java Interface for Solr) to add cores in DataStax Enterprise.