Krishna, Flume is often configured in a scale-out, tiered architecture, and HDFS is a scale-out distributed file system, so the number of data sources is theoretically unbounded, although the limits of physics still apply at any given point. :)
I have seen production Flume deployments with thousands of data sources going into a central HDFS store. Regards, Mike On Thu, Mar 7, 2013 at 1:12 AM, Krishna Kumar < [email protected]> wrote: > Dear Team, > > Please provide me the clarification of the below question > > " How many Data sources FLUME can reliably monitor and load into > centralized storage like HDFS? " > > > Thanks , > T V M Krishna Kumar >
