1

I have a HANA Express Server with version 2 SP 01 and a Hadoop Cluster with Hortonworks Data Platform 2.6. On the Hadoop cluster I installed Vora 1.4 and SparkController 2.1.1.

I saw the following video, which explains the changes from one HANA version to the other: https://www.youtube.com/watch?v=-B9LUjt8x2s&list=PLkzo92owKnVxweu0HK_3QjCfHiMn0jIcA%E2%88%88dex=66

In SPS 10 the Spark Controller seems to be the actual connector for a HANA / Hadoop interaction.

However, could it be, that Vora is the actual standard solution for connecting Hadoop and HANA and the Spark Controller isn't needed anymore in my environment? Or has the Spark Controller some other advantages / functions against Vora?

D. Müller
  • 3,336
  • 4
  • 36
  • 84

1 Answers1

1

Answer of Jason Hinsperger for my other question in the HANA community (https://answers.sap.com/questions/252517/difference-between-spark-controller-and-sap-vora.html):

The Spark Controller is only required if you are connecting to Hadoop from Hana and are not using SAP Vora.

SAP Vora is an independant, distributed in-memory analytics layer for the data lake, and one of its features is that it provides high performance bi-directional access to Hana (eg. you can query Hadoop data from Hana through Vora, and you can query Hana data from Vora running on Hadoop). Vora also has a bunch of other benefits as well related to analyzing data lakes (enhanced functionality, performance, modelling, data specific query engines, etc...) which is covered in the docs and various marketing materials.

HTH,

--Jason

Community
  • 1
  • 1
D. Müller
  • 3,336
  • 4
  • 36
  • 84