- The trial version of TabbyDB will be valid for approx 3 months
- The maximum number of executors that can be spawned is limited to 8
There are two ways to test out TabbyDB
Download and install TabbyDB version 4.0.1 as a fresh install, the way its done for apache spark.
This install is 100% compatible with Spark 4.0.1 .
OR
Convert existing Apache Spark 4.0.1 release installed on your machine to TabbyDB
- Download the tar ball containing 8 TabbyDB jars.
- From your existing spark 4.0.1 install location i.e <SPARK-HOME>/jars, remove the following jars and put it in a separate backup location
- spark-catalyst_2.13-4.0.1.jar
- spark-connect_2.13-4.0.1.jar
- spark-hive_2.13-4.0.1.jar
- spark-sql-api_2.13-4.0.1.jar
- spark-common-utils_2.13-4.0.1.jar
- spark-core_2.13-4.0.1.jar
- spark-repl_2.13-4.0.1.jar
- spark-sql_2.13-4.0.1.jar
3. Unzip the downloaded tabbydb-jars.tar.gz and add the 8 jars it in the location <SPARK-HOME>/jars
The downloaded TabbyDB jars are compatible with Spark 4.0.1 and in most cases, above steps are sufficient.
In case you are using Spark Connect , then you also need to download the TabbyDB's connect client.
Remove any conf property, if present, which excludes any rule of spark
Run your complex use cases and check for performance gains!
If you have any questions/issues please contact at asif.shahid@kwikquery.com
