回复：Jobmanager重启，cannot set up a jobmanager

[绘梦飘雪]

hdfs上 ha storage 目录还在，但里的文件没了，作业占用的资源还在并没有释放，访问flinkui 报service temporarily 
unavailable due to an ongoing leader election



---原始邮件---
发件人: "Yang Wang"

Jobmanager重启，cannot set up a jobmanager

[绘梦飘雪]

jobmanager重启时会org.apache.flink.runtime.client.jobexecutionexception could not 
set up jobmanager
cannot set up the user code libraries file does not exist 
/flink/recovery/appid/blob/job***
到hdfs上看，果然对应文件不存在，请问这个是什么原因造成的

关于DataStreamUtils.reinterpretasKeyedStream的使用

[绘梦飘雪]

hi
  我有这样一个场景，以多个相同的key.做keyby,
DataStream resStream =  demoStream.keyBy(groupKeys)
            .flatMap(new MyFlatmapFunction())
            .keyBy(groupKeys)
            .process(new MyProcessFunction())
            .keyBy(groupKeys)
            .timeWindow(Time.seconds(1))
            .aggregate(new MyAggFunction())
            .keyBy(groupKeys)
            .timeWindow(Time.seconds(1))
            .process(new MyKeyProcessFunction());

我想通过DataStreamUtils.reinterpretasKeyedStream来改写这个流程，我自己尝试这样写下，，
int[] groupKeys = new int[]{0,2,3};

DataStream proStream = DataStreamUtils.reinterpretAsKeyedStream(demoStream, new 
MyKeySelector2(groupKeys) )  //  MyKeySelector2 自己实现keySelector 
.flatMap(new MyFlatmapFunction())
我这样写发现数据流经过flatmap后并不是返回一个keyedstream
,请问是哪里有问题吗？请问该如何使用DataStreamUtils.reinterpretasKeyedStream