报错日志:
org.apache.flink.client.program.ProgramInvocationException: Could not retrieve the execution result. (JobID: 26d1ff5f9e4a48ae6e067e207dfbaf67)
…
Caused by: org.apache.flink.runtime.client.JobSubmissionException: Failed to submit JobGraph.
…
Caused by: org.apache.flink.runtime.rest.util.RestClientException: [Internal server error., <Exception on server side:
akka.pattern.AskTimeoutException: Ask timed out on [Actor[akka://flink/user/dispatcher#931521960]] after [10000 ms]. Sender[null] sent message of type “org.apache.flink.runtime.rpc.messages.LocalFencedMessage”.
akka是flink实现client,jobManager,TaskManager之间的网络通信.
可能报akka ask timeOut的原因
1.可能是由于计算机运行缓慢或网络拥塞引起的
2.资源过少导致 JVM 未能及时响应
解决方案:
1(推荐).当前值:akka.ask.timeout 10s(默认) 修改为100s
web.timeout 10000ms(默认) 修改为300000ms
2.修改JDK版本 当前版本:1.8.0_151 修改版本1.8.0_191
3.当前32个slot分享922MB内存,当前物理主机内存为128G,可以增大内存的分配,增加响应速度