why on mapred-site.xml hadoop 2 there is an mapreduce.jobtracker.address property -
yarn hadoop second generation not use jobtracker daemon anymore, , substitute resource manager. why, on mapred-site.xml hadoop 2 there mapreduce.jobtracker.address property? in order running hadoop mapreduce application eclipse , there plugin eclipse yarn , because find plugins spesified jobtracker. in advance. hadoop 2 compatible run on yarn cluster mr cluster . just in case 1 needs run hadoop 2 compiled code on non-yarn cluster 1 has specify jt url:port . hadoop 2 in binary compatible on both mr/yarn cluster .