1

I recently installed Hadoop 1.0.3 from the .deb provided by Apache. The package installed correctly, but there seem to be two directories that have Hadoop-related files: /usr/share/hadoop has jars and the site configuration files, while /etc/hadoop has hadoop-env.sh and some other XMLs; all of the Hadoop literature I've read thus far seems to assume there is only a single Hadoop install directory; could someone please elaborate on these two directories and their respective purposes?

Thanks.

Improve this question

1 Answer 1

Reset to default
2

/usr/share is a platform-independent shared data directory (and hadoop is in java...) and /etc/hadoop is mostly host specific system and application configuration files.

Then, to configure hadoop, just use the following command and answer yes, yes, yes... !

sudo ./usr/sbin/hadoop-setup-single-node.sh

Then

sudo ./usr/sbin/hadoop-validate-setup.sh --user=hdfs

Complete tutorial for single node

Improve this answer
3
  • Hi Yohann, Thanks for your answer. Would it be accurate to say that /etc/hadoop is $HADOOP_HOME, and if so, why does it not include core-site.xml, mapred-site.xml, and hdfs-site.xml? Should I create a conf directory and copy /usr/share/hadoop/templates/conf/*-site.xml to /etc/hadoop/conf/ ?
    – ILikeFood
    May 22, 2012 at 19:50
  • See the edit in my post.
    – Yohann
    May 22, 2012 at 20:01
  • very helpful for me as well!
    – Alex Gordon
    May 3, 2013 at 22:49

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .