Amazon EMR
Developer Guide

Updates for Impala 1.2.4

The following updates are relevant for Impala 1.2.4:

  • Performance improvements on metadata loading and synchronization on Impala startup. You can run queries before the loading is finished (and the query will wait until the metadata for that table is loaded if it's necessary).

  • INVALIDATE METADATA was modified to accept argument, table_name, to load metadata for a specific table created by Hive. Conversely, If the table has been dropped by Hive, Impala will know that the table is gone.

  • You can set the parallelism of catalogd's metadata loading using --load_catalog_in_background or --num_metadata_loading_threads


The following features were added with Impala 1.2.3 and 1.2.2 and are available with this release.

Update for Impala 1.2.3

  • Notable bug fix: compatibility with Parquet files generated outside of Impala.

Updates for Impala 1.2.2

  • Changes to Join order. Impala will automatically optimize a join query to minimize disk I/O and network traffic, instead of making a user order the join in a specific fashion.

  • Use the STRAIGHT_JOIN operator to bypass Impala's automatic optimization of table order in a join query.

  • Find table and column information with COMPUTE STATS.

  • Use the CROSS JOIN operator in a SELECT statement for queries needing Cartesian products.

  • New clauses for ALTER TABLE.

  • LDAP authentication for JDBS and ODBC drivers.

  • Numeric and conditional functions can return SMALLINT, FLOAT, and other smaller numerical types.

For more information, see: