Hi, I'm confused with storm cluster size.I have 3 topologies where they are pulling data from different kafka-sources.I deployed them on 12 supervisor with 1 Nimbus of 2gb ram,1 core machines each, my problem is with the processing speed.They are processing the message with 10 mins delay.
Is it the problem of hardware configurations ? In storm link <http://storm.apache.org/about/scalable.html> they suggested 24 memory with 2 core machine for production-cluster. Does it mean I should go with 24 memory machine vm's? I.e., now I'm having 2gb ram machines as per test. if I want to move them to production then I need to increase each vm ram size of 2gb to 24gb ? Is it a right assumption? or do I need to check the topology code again ? *Best regards,* *K.Sai Dilip Reddy.*