- Add a variable to define the
spark
user's shell. - Add a variable to associate existing groups with
spark
user. - Add variables for customizable Spark work and temporary directories.
- Add backward compatible support with Ansible 2.0.
- Ensure that example uses YAML role specification format.
- Add configurable Spark user and group.
- Add support for a custom installation directory.
- Add support for settng
spark-defaults.conf
.
- Use Apache as the source for Spark vs. Cloudera. This change breaks backward compatibility with roles < 1.0.0.
- Bump Spark version to
1.3.0
. - Add
spark_cloudera_distribution
variable to determine which package distribution channel is used.
- Add ability to set arbitrary environment variables in
spark-env.sh
- Fix Hadoop configuration file reference in
spark-env.sh
.
- Bump Spark version to
1.2.0+cdh5.3.0+364-1.cdh5.3.0.p0.36~trusty-cdh5.3.0
.
- Initial release.