Spark can be a consumer and a producer from the Kafka point of view. You can create a kafka client in Spark that registers to a topic and reads the feeds, and you can process data in Spark and generate a producer that sends that data into a topic. So, Spark lies next to Kafka and you can use Kafka as a channel to collect and send the data.
That’s what I am doing, at least. > On 22 Mar 2017, at 08:08, Adaryl Wakefield <adaryl.wakefi...@hotmail.com> > wrote: > > I’m a little confused on how to use Kafka and Spark together. Where exactly > does Spark lie in the architecture? Does it sit on the other side of the > Kafka producer? Does it feed the consumer? Does it pull from the consumer? > > Adaryl "Bob" Wakefield, MBA > Principal > Mass Street Analytics, LLC > 913.938.6685 > www.massstreet.net <http://www.massstreet.net/> > www.linkedin.com/in/bobwakefieldmba > <http://www.linkedin.com/in/bobwakefieldmba> > Twitter: @BobLovesData
signature.asc
Description: Message signed with OpenPGP