You can specify the query engine for Treasure Data. This setting specifies the query engine to be selected for queries in the TD Console. Your choices are:
- Presto for ad hoc and shorter batch workloads. Presto provides low-latency SQL access to the data set.
- Hive for large or complex batch workloads. Hive is a MapReduce-based SQL engine. This engine is really powerful when you do large data processing and heavy JOINs. Often used for ETL or sessionization.
Treasure Data is dedicated to making the transition between query engine versions as smooth as possible for all of our customers. To facilitate this, in some cases, we get out of sync with the latest open source versions of the engines. It also means that sometimes, features that are generally available in the newest versions of the engines are still being tested for our platform and are not available for our customers. The best way to take advantage of the performance optimizations that we continue to rollout is to adopt the newest versions of the query engines that we make available.
About TD Query Engine Release Numbering
- TD Presto release numbering uses the open source Presto release numbers.
- TD Hive release numbering is unique to Treasure Data.
|Presto Query Engine Support||Hive Query Engine Support|
|Presto 0.205||Hive 0.13 is based on open source Apache Hive 0.13.|
|Hive 2019.1 is based on open source Apache Hive 2.3.2.|
About Hive Version Features
|Hive 0.13||Hive 2019.1|
To specify the query engine for a specific query
You can configure the query engine for Treasure Data queries. In some cases, choosing one query engine over the other might optimize performance or manage your data accuracy better. The default engine for your account is shown in bold.
- Navigate to the TD Console > Data Workbench > Queries.
- Select an existing query or create a new query.
- In the Type field, specify the query engine and version you want to use for the query.
- Optionally edit or type your new query.
- Save your choice.
To specify your account-wide query engine
You can configure the account-wide default query engine for Treasure Data. All new queries will have the engine specified as the default query engine.
- Select the Control Panel icon.
- Select Account Info.
- Select the Edit icon.
- Validate your account information.
- For Default Query Engine, make one of the following choices:
- Save your choice.