spark集群详细搭建过程及遇到的问题解决（二）

（二）配置ssh无密码访问集群机器

master节点

root@master:/home# su - spark
spark@master:~$
spark@master:~$ ssh-keygen -t rsa
#一直enter键

spark@master:~$ cd .ssh/
spark@master:~/.ssh$ ls
id_rsa  id_rsa.pub
spark@master:~/.ssh$  cat id_rsa.pub > authorized_keys
spark@master:~/.ssh$ scp spark@master:~/.ssh/id_rsa.pub ./master_rsa.pub

spark@master:~/.ssh$ ls
authorized_keys  id_rsa  id_rsa.pub  known_hosts  master_rsa.pub
spark@master:~/.ssh$ cat master_rsa.pub >>authorized_keys

worker1节点
与master节点执行过程相同

worker2节点与master节点执行过程相同

执行完上述操作之后：
master节点：

spark@master:~/.ssh$ scp spark@worker1:~/.ssh/id_rsa.pub ./worker1_rsa.pub

注意是在./ssh目录下

spark@master:~/.ssh$ cat worker1_rsa.pub >>authorized_keys
spark@master:~/.ssh$ scp spark@worker2:~/.ssh/id_rsa.pub ./worker2_rsa.pub

spark@master:~/.ssh$ cat worker2_rsa.pub >>authorized_keys

worker1节点

spark@worker1:~/.ssh$ scp spark@worker2:~/.ssh/id_rsa.pub ./worker2_rsa.pub
cat worker2_rsa.pub >>authorized_keys

worker2节点

spark@worker2:~/.ssh$ scp spark@worker1:~/.ssh/id_rsa.pub ./worker1_rsa.pub
cat worker1_rsa.pub >>authorized_keys

验证是否配置成功

从master节点分别登陆自身、worker1节点与worker2节点（可能第一次需要密码、退出后，第二次重新ssh则不需要密码）

spark@master:~/.ssh$ ssh master

spark@master:~/.ssh$ ssh worker1

spark@master:~/.ssh$ ssh worker2

从worker1节点登陆自身、master节点与worker2节点（可能第一次需要密码、退出后，第二次重新ssh则不需要密码）

spark@worker1:~/.ssh$ ssh master

spark@worker1:~/.ssh$ ssh worker1

spark@worker1:~/.ssh$ ssh worker2

从worker2节点登陆自身、master节点与worker1节点（可能第一次需要密码、退出后，第二次重新ssh则不需要密码）

spark@worker2:~/.ssh$ ssh master

spark@worker2:~/.ssh$ ssh worker1

spark@worker2:~/.ssh$ ssh worker2

注意，登陆到其他节点后，一定要记得退出
下文中，将开始安装hadoop。。。