Комментарии:
Should I learn Hadoop now or can directly go to spark?
Ответитьi think for machine learning spark is best what's your view on this?
ОтветитьGreat Explanation
ОтветитьHadoop is just 2 components? well, what about yarn?
ОтветитьCould you make a video on which certification is recommended for spark?
Ответитьcode???
Ответитьwith all those cloud vendors offering large storage, is hadoop for organizations dying??
Ответить@Thomas Henson. Please considering testing out some smart sound or smart volume software. I am having to chase your speaking volume.. One second it is too loud and will annoy the neighbors or even me. And the next minute I am backing up the video to hear what I missed. Sorry to be a complainer but I love your videos. You seem friendly and your information is thorough and easy to understand. (If I can hear it.) ;)
Ответитьwhere is the link to the blogpost?
Ответитьwith all those cloud vendors offering large storage, is hadoop for organizations dying?
ОтветитьThe main problem with Hadoop is that it is not optimized to handle iterative data processing. The HDFS system becomes overburden with constant assesses as the output of reduce becomes input to Map over and over until the final convergences takes place. Also, the data processing is Acyclic and you cannot interact with each stage of processing. Even though you can do iterative processing on Hadoop it is not very efficient; Hence Spark !!
ОтветитьTHANKS MAN THAT WAS REALLY HELPFUL
Ответитьty
Ответить