Search code examples
akka-streamalpakka

Alpakka Cassandra Flow Usage

I am reading the documentation of Alpakka Cassandra here

It makes it very easy to use Cassandra as a source and a sink. but what about flow usage.

By flow usage I mean is that I am not using Cassandra as a source or a sink. but to Lookup Data.

is that possible using Alpakka? or should I have to write the Cassandra jdbc code in a flow myself?

Solution

  • 1) Sink. If you check Alpakka's source code, you'll find that the Sink is constructed as follows

    Flow[T]
      .mapAsyncUnordered(parallelism)(t ⇒ session.executeAsync(statementBinder(t, statement)).asScala())
      .toMat(Sink.ignore)(Keep.right)

    if you only need a passing flow, you can always trim out the Sink.ignore part, and you'll have

    Flow[T]
          .mapAsyncUnordered(parallelism)(t ⇒ session.executeAsync(statementBinder(t, statement)).asScala())

    You will only need to expose the Guava futures converter, which is currently package private in Alpakka.

    2) Source. You can always obtain a Flow from a Source by doing .flatMapConcat(x => CassandraSource(...))