We have a java Spark streaming job that basically read messages from Kafka and writes outputs to Kafka, Hive or HBase.
We want to be compliant with both HDP2.6 and HDP3.1.5. Is it possible to build a single software that can run either on HDP2.6 or HDP 3.1.5?
What is the best approach for this?
HDP2.6 uses java 7 and HDP 3.1.5 uses java 8.
If you want to write a program that works for both versions,
you can use the compile option to run on java 7.
Thank you but both version support java 8, no problem on this side.
See : https://supportmatrix.hortonworks.com/
The main problems are the new HBase version with HDP 3.1.5 and the new Hive version with the addition of the Hive Warehouse Connector on HDP 3.1.5.