[英]How to configure java home on flink job on yarn
I submit a job to the hadoop cluster on the flink client, do I need to configure the java home path in flink conf?我在flink客户端提交作业到hadoop集群,是否需要在flink conf中配置java home路径?
If you need to configure,如果需要配置,
Should I configure the java home of the client machine or the java home of the hadoop cluster?我应该配置客户端机器的 java home 还是 hadoop 集群的 java home?
When I did not configure java home, I submitted the job error as follows当我没有配置java home时,提交作业报错如下
LogType:jobmanager.err
Log Upload Time:Fri Jan 22 17:27:25 -0800 2021
LogLength:160
Log Contents:
Unrecognized VM option 'MaxMetaspaceSize=268435456'
Error: Could not create the Java Virtual Machine.
Error: A fatal exception has occurred. Program will exit.
According the output you posted, you problem is not java_path (yet!) The line根据您发布的 output ,您的问题不是 java_path (尚未!)行
Unrecognized VM option 'MaxMetaspaceSize=268435456'无法识别的 VM 选项“MaxMetaspaceSize=268435456”
You need to remove MaxMetaspaceSize from configuration or replace it with: -XX:MaxMetaspaceSize您需要从配置中删除 MaxMetaspaceSize 或将其替换为:-XX:MaxMetaspaceSize
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.