Build and Runtime Dependencies

Starting with version 4.0 Elephant-Bird defines majority of its dependencies in maven provided scope. As a result these dependencies are not trasitively pulled by application that depend on Elephant-Bird modules. Almost all the time, projects define their dependencies for Hadoop, Pig, etc and end up exluding these libraries from Elephant-Bird transitive dependencies. These libraries are normally provided by by execution environment. E.g. pig command sets up all the libraries for Pig.

The following are some of the libraries required by Elephant-Bird moduled at runtime. pom.xml lists actual versions, though Elephant-bird does not have strict dependency on specific versions.

1. hadooop-client
2. hadoop-lzo
3. pig (for elephant-bird-pig)
4. cascading (for elephant-bird-cascading)
5. hive (for elephant-bird-hive)
6. hcatalog (for elephant-bird-hive)