Our company is using hdp clusters.
My boss asked me to configure hive on Spark.
I know hortonworks dosen't support hive on Spark.
But I can't convince my boss.
So can I have any alternative methods to realize it ?
Such as building the original spark source code and installing it on hdp? or any other methods else.
Really hope anyone can help . Thanks very very much!!
I quite surprised by your analogy, what exactly do you want to do. I have participated in projects where spark is having used against HDP cluster. Zeppelin note but is used to develop and run spark and hive.
What is your use case if you can explain what you intend to achieve then members can contribute.
HDP supports Hive On Spark. You just set the value ‘spark’ of hive engine in shell or query. But I strongly recommend to use Hive On Tez in case of HDP cluster.