HI All,

I am wondering if it makes sense to have two receivers inside my Spark
Client program?

The use case is as follows.

1) We have to support a feed from Kafka so this will be a direct receiver
#1. We need to perform batch inserts from kafka feed to Cassandra.

2) an gRPC receiver where we get a RPC request or HTTP request (since grpc
uses HTTP2) to retrieve a record  or perform a table scan or select a good
chunk of records from Cassandra. so this is more like a microservice with
spark streaming layer underneath where some of the requests such as batch
reads and batch inserts will go through spark.

Any feedback or thoughts will be great!


Reply via email to