Hive 2 Upgrade Overview
Hive 0.13 was a core query engine for Treasure Data. Treasure Data has decided to move forward with support for Hive 2. If you are still on Hive 0.13, it is highly recommended that you migrate the queries that you may have on TD to accommodate the new syntax supported by Hive 2.
Hive 2 provides:
- Ongoing open source community focus on the Hive 2.x (and very similar Hive 3.x) code lines.
- Substantial enhancements, including: Bug fixes and stability improvements
- Support for improvements in the underlying Hadoop environment
- Support of Tez
- Ongoing query optimization performance Improvements
- Resource management
- Introduction of ANSI SQL compliance for direct portability between Hive and Presto SQL dialects in the TD environment
Advantages of Tez include:
- Execution Performance
- Performance gains over Map Reduce
- Optimal resource management
- Plan reconfiguration at runtime
- Dynamic physical data flow decisions
Rewriting existing Hive 0.13 workloads to run in Hive 2 will be required to take advantage of most of the enhancements. TD Hive 0.13 will continue to be supported for the time being, but will be phased out soon.
When re-writing queries for Hive 2, we recommend that you use TD from the command line to EXPLAIN the query.