Use this to build a kafka RPM.

This project is laid out for use with rpmbuild.  Let's suppose you've cloned it
as ~/kafka-rpm.  Then to build the RPM, you'd do:

    ln -s kafka-rpm rpmbuild
    spectool -g -R rpmbuild/SPECS/kafka.spec
    rpmbuild -bb rpmbuild/SPECS/kafka.spec

The 'spectool' line will download the sources for kafka and the binaries for
gradle.  You could alternatively download them by hand and put them in the
SOURCES directory.  The rpmbuild line builds kafka and creates the RPM
containing just the files needed for running kafka, which will be placed in a
subdirectory of the RPMS directory.

The result of installing the RPM is similar to the kafka binary distribution,
but installed.  The jar and support files are placed in /opt/kafka, Two users
are added, kafka and zookeeper.  All the logging goes to /var/log/kafka and
/var/log/zookeeper.  The data files are placed in /var/lib/kafka and
/var/lib/zookeeper.  All of the config files mentioned in the kafka quickstart
are located in /opt/kafka/config-sample.  Config files are placed in
/opt/kafka/config to allow the supplied init scripts to start zookeeper and
kafka on their default ports, but you will need to change these configs for a
multi-server production deployment.  Once installed zookeeper and kafka will be
started on boot; run chkconfig if you don't want that.

NOTE: the spec file requires java 1.8 for security reasons.  Kafka will build
with 1.7 if you want to figure out if you have a 'safe' version of 1.7,
or don't care.

Updating the kafka version should be a straightforward matter of changing
the version number...until they change their build process again.