I had the same problem and I also solve it adding the user to all controler nodes. Run a script for add them from on_to_all ______________________________________________________________________________________________ #!/bin/bash # Linux/UNIX box with ssh key based login SERVERS=/root/hadoop_hosts # SSH User name USR="root" # Email SUBJECT="Server user login report" EMAIL="your_e-mail@here" EMAILMESSAGE="/tmp/sshpool_`date +%Y%m%d-%H:%M`.txt" # create new file >$EMAILMESSAGE # connect each host and pull up user listing for host in `cat $SERVERS` do echo "--------------------------------" >>$EMAILMESSAGE echo "* HOST: $host " >>$EMAILMESSAGE echo "--------------------------------" >>$EMAILMESSAGE ###ssh $USR@$host w >> $EMAILMESSAGE ssh -tq -o "BatchMode yes" $USR@$host $1 >> $EMAILMESSAGE done # send an email using /bin/mail ######/bin/mailx -s "$SUBJECT" "$EMAIL" < $EMAILMESSAGE echo ">>>>>>>>>>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<<<<<<<<<<<<" echo ">>> check the output file " $EMAILMESSAGE _________________________________________________________________________________________ put DNS servers names into /root/hadoop_hosts Also in linux there is a good command called pssh to run comands in parallel in computer clusters 🙂 ... View more

I have same problem, how did you solve it? also yarn leave no log information when I do yarn log -applicationid application_1489695782031_0588 Error: Could not find or load main class log ... View more