It had been seven years since I last installed a multi-cluster Apache Hadoop node, so it was naturally a bit nostalgic for me to implement one today. In the times since I started using linux-based operating systems, I've come to know it quite well.

I noticed something perhaps very common with the Hadoop installation guides that I couldn't have noticed seven years ago: they all asked you to modify the PATH variable in either ~/.bashrc, /etc/environment, or both, to set different points for the hadoop bin and sbin executables. As I went on, I noticed that from the extracted folder, the config files are ~/hadoop/etc/hadoop, which seems to be a strange place for the config files. Why not simply name it, hadoop/config like most other self-contained zip-drop packages. Unless, of course, the contents of the Hadoop configuration file are designed to be installed to /etc/hadoop, meaning the entire contents of the tarball you download are meant to be extracted directly to root directory /.

While it obviously works fine like I have it set up (in the /home/hduser/hadoop folder), I was wondering how others have configured their Hadoop installations, and if anyone else thinks it could be beneficial to write a tutorial that places Hadoop's bin and sbin folders in /usr/bin and /usr/sbin. And, of course, their configs in /etc/hadoop.

Does anyone think writing an automated installer/add-nodes script would be beneficial?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.