In my last three weeks I have been deployed to a different business unit, here I have been working on Hadoop distributed file system HDFS. Installation of Hadoop cluster and configuration process is roughly as follows, initially we have to select a machine as a Master Hadoop configuration on the Master host user to install SSH server, install the Java environment Hadoop is installed on the Master host, and complete the configuration and also configuring hadoop users on other hosts, install SSH server, install the Java environment Copy Hadoop directory on the Master host to another host Open.
It’s is strongly recommended for novice first installed in accordance with the above tutorial on the Master host once pseudo-distributed environment of Hadoop, after all, usually written procedures, or will be carried out in pseudo-distributed debugging environment. Pseudo be distributed to normal use, and then continue to browse this tutorial subsequent cluster configuration.
Here, I have used the two hosts to build a cluster, the host name and IP address correspond to the following: Which hosts the first selected for a Master (for example, I chose this ip, and then in the host's /etc /hostname , modify the machine named Master, the other hosts command Slave1,…show more content… After configuring the Hadoop file copied to the Master on each node (although direct use scp copy can also run correctly, but it will be different, such as symbolic links scp past after a bit different. So the first pack and then copy Comparison safe). Because before there ran a pseudo-distributed mode recommended for temporary files before switching to the cluster pattern must
