All,
Any thoughts? I can run another couple of experiments to try to narrow the
problem. The total data volume in the repartition is around 60GB / batch.
Regards,
Bryan Jeffrey
On Tue, Dec 13, 2016 at 12:11 PM, Bryan Jeffrey
wrote:
> Hello.
>
> I have a current
Hello.
I have a current Spark 1.6.1 application that I am working to modify. The
flow of the application looks something like the following:
(Kafka) --> (Direct Stream Receiver) --> (Repartition) -->
(Extract/Schemitization Logic w/ RangePartitioner) --> Several Output
Operations
In the