问题
I am trying to run a hadoop 2.2.0 mapreduce job on my local single node cluster installed by following this tutorial: http://codesfusion.blogspot.co.at/2013/10/setup-hadoop-2x-220-on-ubuntu.html?m=1
Though on the server side the following exception is thrown:
org.apache.hadoop.ipc.RpcNoSuchProtocolException: Unknown protocol: org.apache.hadoop.yarn.api.ApplicationClientProtocolPB
at org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ProtoBufRpcInvoker.getProtocolImpl(ProtobufRpcEngine.java:527)
at org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ProtoBufRpcInvoker.call(ProtobufRpcEngine.java:566)
at org.apache.hadoop.ipc.RPC$Server.call(RPC.java:928)
at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2048)
at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2044)
at java.security.AccessController.doPrivileged(Native Method)
at javax.security.auth.Subject.doAs(Subject.java:415)
at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1491)
at org.apache.hadoop.ipc.Server$Handler.run(Server.java:2042)
Is there a way for me to configure Protobuf RPC to be available on server side? Do I need the hadoop native libraries for this? Or can I switch somehow on the client side to Writables/Avro RPC?
回答1:
Ok, found the reason, I connected to the wrong port for the yarn resourcemanager. The correct configuration is: yarn.resourcemanager.address=localhost:8032
回答2:
In my case, I was getting the same error in the logs when there was not enough memory between the Application Master and the YARN containers. Reduced the yarn.app.mapreduce.am.resource.mb property and it worked on my single node installation.
来源：https://stackoverflow.com/questions/20163657/protobuf-rpc-not-available-on-hadoop-2-2-0-single-node-server