1

I want reduce load on a single server which is currently hosting EventServer, PredictionServer and Training. Obviously, this will not be scalable.

I want to push the SPARK training job on an EMR cluster which starts on demand, so that our main server is not consumed.

pio train -- --master=spark://master-emr-ip:7077

It could not find that remote. Security groups has been set rightly.

ANKIT HALDAR
  • 1,365
  • 1
  • 10
  • 17
  • Have a look at this and rethink if you still want to do it: https://lists.apache.org/thread.html/%3CCAMBzQgwCLj9QfmmB2uwQyUhprmD1cuAvvn4d2mENO-G2t7HcYQ@mail.gmail.com%3E – Chris W. Jun 07 '18 at 11:55
  • Thanks for your comment as we hoped to try the same with EMR – Chris W. Jun 07 '18 at 16:30

0 Answers0