Skip to content

Latest commit

 

History

History
39 lines (27 loc) · 801 Bytes

README.md

File metadata and controls

39 lines (27 loc) · 801 Bytes
Raw

spark

Spark cluster computing system for big data

Build the roll

Checkout a roll distribution from github:

git clone https://github.com/pragmagrid/spark
cd spark
./bootstrap.sh

The above will download spark tarball v 2.3.0 and hadoop v 2.7 To update to a different version, edit src/spark/version.mk and rerun boostrap.sh

Downloads archive is https://spark.apache.org/downloads.html

Build the roll

make roll

Resulting ISO file spark-.iso is the toll that can be added to a frontend

Add the roll

To add the roll to a frontend

# rocks add roll spark-7.0-0.x86_64.disk1.iso
# rocks enable roll spark
# (cd /export/rocks/install; rocks create distro)
# yum clean all
# rocks run roll spark > add-spark.sh
# bash add-spark.sh