parallel processing - Apache Spark vs Akka -
could please tell me difference between apache spark , akka, know both frameworks meant programme distributed , parallel computations, yet don't see link or difference between them.
moreover, use cases suitable each of them.
apache spark built on akka.
akka general purpose framework create reactive, distributed, parallel , resilient concurrent applications in scala or java. akka uses actor model hide thread-related code , gives simple , helpful interfaces implement scalable , fault-tolerant system easily. example akka real-time application consumes , process data coming mobile phones , sends them kind of storage.
apache spark (not spark streaming) framework process batch data using generalized version of map-reduce algorithm. example apache spark calculation of metrics of stored data better insight of data. data gets loaded , processed on demand.
apache spark streaming able perform similar actions , functions on near real-time small batches of data same way if data stored.
update april 2016
from apache spark 1.6.0, apache spark no longer relying on akka communication between nodes. @eugenemi comment.
Comments
Post a Comment