What's New in CDH 5.14.x

• New Boolean operators: IS TRUE, IS NOT TRUE, IS FALSE, IS NOT FALSE, IS UNKNOWN, and IS NOT UNKNOWN. These operators always return TRUE or FALSE, even if the left-hand expression evaluates to NULL. IS [NOT] TRUE and IS [NOT] FALSE are NULL-safe equivalents for equality and inequality comparisons. IS [NOT] UNKNOWN is identical to IS [NOT] NULL.

• The new built-in function DATE_TRUNC() lets you truncate a TIMESTAMP value to a given precision. For example, calling date_trunc('hour',ts) truncates ts to the beginning of the corresponding hour, with all minutes, seconds, milliseconds, and so on set to zero.

• The impala-shell command now accepts a command-line argument --query_option=option=value to set a query option for that shell session. You can specify multiple query options by supplying multiple instances of --query_option in the impala-shell invocation.

  Configuration files for impala-shell, such as .impalarc, can also contain a section beginning with [impala.query_options], followed by lines of the form query_option=value. These query options take effect by default for all impala-shell sessions that use the specified configuration file.

• The SET statement, when run without any argument, now divides its output into groups corresponding to "regular", "advanced", "development" (unfinished, not ready for general use), and "deprecated" (subject to removal). By default, the "development" and "deprecated" option groups are not shown in SET output. Issue the statement SET ALL to see the options and values from all groups.

• Impala optimizes some join queries against Kudu tables by constructing runtime filters representing the minimum and maximum matching values from the join columns. The min-max information is "pushed down" to Kudu so that Kudu can prune unnecessary rows before sending the result set back to Impala.

• The allowed range for the BATCH_SIZE query option is now 0 to 65536. Formerly, the upper bound was undefined.

• Arithmetic expressions involving both DECIMAL and FLOAT or DOUBLE arguments now produce results of DECIMAL type instead of DOUBLE.

Starting with Apache Kudu 1.5.0 / CDH 5.13.x, Kudu has been fully integrated into CDH. Kudu now ships as part of the CDH parcel and packages. The documentation for Kudu has also been incorporated into the Cloudera Enterprise documentation here.

For a complete list of new features and changes introduced in Kudu (in CDH 5.14), see What's New in Apache Kudu.

In previous versions, the Lily HBase Indexer service only supported Sentry policy files for authorization. It now supports the Sentry Service. For more information, see Configuring Lily HBase Indexer Security.