flink集群安装后启动报错

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

flink集群安装后启动报错

赫里斯汀的树
大佬们你们好:
     我在Linux服务器上安装flink集群(3台机器)时,版本1.9.3 
         在机器1上启动集群执行 ./start-cluster.sh 报错 Name or service not knownname ce-hjjcgl-svr-02(机器2),机器1和机器3上的正常启动;
         [appuser@ce-hjjcgl-svr-01 bin]$ ./start-cluster.sh
         Starting cluster.
         Starting standalonesession daemon on host ce-hjjcgl-svr-01.
         : Name or service not knownname ce-hjjcgl-svr-02
         Starting taskexecutor daemon on host ce-hjjcgl-svr-03.


         
         
         
         在机器2上启动./start-cluster.sh,也报这个错:
         [appuser@ce-hjjcgl-svr-02 bin]$ ./start-cluster.sh
        Starting cluster.
        [INFO] 1 instance(s) of standalonesession are already running on ce-hjjcgl-svr-02.
        Starting standalonesession daemon on host ce-hjjcgl-svr-02.
        : Name or service not knownname ce-hjjcgl-svr-02
         


         
         集群之间已实现ssh免密登录,三台机器的hosts文件如下:
         127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
        ::1         localhost localhost.localdomain localhost6 localhost6.localdomain6
        10.212.139.219 ce-hjjcgl-svr-01
        10.212.139.220 ce-hjjcgl-svr-02
        10.212.139.221 ce-hjjcgl-svr-03
       
        三台机器的masters文件:
        ce-hjjcgl-svr-01:8081
       
        slaves文件:
        ce-hjjcgl-svr-02
        ce-hjjcgl-svr-03
       
       
        机器2的flink-conf.yaml文件部分:
       
        env.java.home: /usr/java/jdk1.8.0_202/


        jobmanager.rpc.address: ce-hjjcgl-svr-02
         
  jobmanager.rpc.port: 6123
       
       
       
        还望大佬们帮忙看看,不胜感激!