Real time + batch processing of data
I am evaluating a distributed execution framework for a combination of real time (like breaking news) and batch data (analytics, recommendations) on largely news like content. Hoping to use a framework that has a single copy of data, common set of resources and allowing for near real time processing of content and as well much batch jobs.
I looked at storm-yarn open sourced at Yahoo and seems to fit the bill. Are there any other alternate choices i could look at? I guess in the workst case, i could use both hadoop and storm independently?