Running hadoop as a linux service

Question

I am neither an expert to linux not to hadoop but I followed the instruction for installation hadoop in distributed mode(Hortonworks distribution and manual) and when I got to section starting the cluster I give a thought why this couldn't be run as a service.

Does anybody have some experience with running hadoop as a linux service? Is it a good or bad idea? Pros and cons? Any hints doing that?

Any help will be appriciated. Thx

Answer 1

是的，您可以运行hadoop作为服务，因为您需要编写小的脚本来设置环境并启动适当的nodemanager ，例如slave服务器上的datanode和nodemanager以及hdfs服务器上的namenode和yarn server上的resourcemanager等，然后将脚本保存到/ etc / init.d /，那么您需要在rcX.d中适当指定创建链接以获取更多update-rc.d