Vora is showing the commitment that SAP is making in the new Big Data world, and has been very well received by our customers as evidenced by the numerous visits the Vora booth recently got at TechEd Las Vegas and being part of Steve Lucas' demo with Bluefin Solutions at TechEd Barcelona event.
So, we have been working hard to get Lumira to connect to Vora and get a nice looking front-end BI tool on top of this new engine. The following instructions explains how to make this happen with the release of Lumira 1.29:
- Install the SparkSQL driver delivered in Lumira 1.29
- We will need to use the Spark drivers as Generic JDBC driver:
- Go to Lumira -> Preferences -> SQL Drivers -> Generic JDBC datasource, click on "Install Drivers"
- Select all *.jar files under C:\Program Files\SAP Lumira\Desktop\utilities\SparkJDBC
- Restart Lumira
- We will need to use the Spark drivers as Generic JDBC driver:
- Connect now to SAP HANA Vora
- New datasource > SQL connectivity > Generic JDBC Parameters
- Connect to your Spark cluster
- JDBC Class: “com.simba.spark.jdbc4.Driver”
- JDBC URL: jdbc:spark://[host]:[port]/default;CatalogSchemaSwitch=0;UseNativeQuery=1
- Querying SAP HANA Vora
- Default queries won't work currently: SELECT * from “SPARK”.”default”.”<table name>” syntax is not compatible with Spark for a couple of reasons (one is syntax as back-tick is the identifier quote character and the other relates to how table metadata is being declared in the catalogs and how they are consumed)
- Just write down the table name directly in the query and those will work. The default query as generated by default by Lumira need to therefore be adjusted. So, in summary:
- select * from "Spark"."default"."TABLE"."sensor1_h_rep" won't work
- select * from sensor1_h_rep works
Curious to hear about your experience using this end-to-end analytical solution from SAP!
Thanks,
Paul
- SAP Managed Tags:
