Closed markkimsal closed 10 years ago
To keep the datanode alive after the namenode -format command I had to source the yarn-env.sh file before formatting the name node
. /usr/local/hadoop/etc/hadoop/yarn-env.sh
I still cannot run the MR example. netstat -ap shows no programs listening to 8032.
Ahh, apparently the second sbin/start-dfs.sh should be sbin/start-yarn.sh. I was confused because running format killed my datanode without the yarn environment variables.
New Testing section of README should be:
# start ssh and hdfs
service sshd start
. /usr/local/hadoop/etc/hadoop/hadoop-env.sh
. /usr/local/hadoop/etc/hadoop/yarn-env.sh
cd $HADOOP_HOME
sbin/start-dfs.sh
# format and create directories
bin/hdfs namenode -format
sbin/start-yarn.sh
bin/hdfs dfs -mkdir -p /user/root
bin/hdfs dfs -put etc/hadoop/ input
# run the mapreduce
bin/hadoop jar share/hadoop/mapreduce/hadoop-mapreduce-examples-2.3.0.jar grep input output 'dfs[a-z.]+'
# check the output
bin/hdfs dfs -cat output/*
Mark thanks for your effort. This work is still in heavy construction. Actually i'm just moving the first section into the image itself. So the image will:
Stay tuned ;)
Running the example steps from README.md, when I get to the MR example I get the following output:
The last message repeats over and over endlessly. (No files end up in HDFS output/ folder either)
My datanode process is not running, stoping and starting hdfs doesn't launch it.