Github user StephanEwen commented on the pull request:
https://github.com/apache/flink/pull/1342#issuecomment-155769960
Robert has a good point. Right now, if some one uses HBase, the HBase
dependency is part of the user program JAR. It would be nice to keep HBase out
of the core JAR - it would be yet another fat dependency with multiple
transitive dependencies.
Once you change the code to reflect-load the HBase classes, one gets the
Kerberos/HBase support as soon as one drops the HBase jars JAR into Flink's lib
folder, or adds them to the Hadoop Classpath env variable. At the same time,
non-HBase users retain a thinner set of dependencies (and possible conflicts).
@nielsbasjes What do you think about that approach?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
