I'm using Apache Flink to process streaming data.
Saying that I have two jar files a.jar and b.jar. My question is if it is possible to use the sink of a.jar as the data source of b.jar.
If it's possible, can you show me a simple example?
If not, what could I do? Must I use something like a Database to store the output of the sink of a.jar and use the Database as the data source of b.jar?
To get exactly-once guarantees, you'll have to implement this decoupling with a stream storage layer that supports transactions and stream replay.