llamasoft opened a new pull request #4244: URL: https://github.com/apache/zeppelin/pull/4244
### What is this PR for? The `bin/zeppelin.sh` startup script doesn't fully load the Hadoop classpaths in the same way that `bin/zeppelin-daemon.sh` does. As a result, anything that depends on the Hadoop classpaths being loaded (e.g. Shiro's KnoxJwtRealm) will fail. ### What type of PR is it? Bug Fix/Improvement ### How should this be tested? * Verify that `hadoop classpath` values are added to Zeppelin's java process's classpath. * Verify that Shiro's KnoxJwtRealm loads correctly when running `USE_HADOOP=true HADOOP_CONF_DIR=/path/to/conf bin/zeppelin.sh` * Verify that the Zeppelin Docker image still works as expected. ### Notes: * The classpath loading logic was copied directly from `bin/zeppelin-daemon.sh`. * The Zeppelin Docker image uses `bin/zeppelin.sh` as its startup script. This PR will make it easier for users to extend the Docker image by adding Hadoop to it. Currently, adding Hadoop requires extensive modifications to get the classpath to load correctly. ### Questions: * `bin/zeppelin-daemon.sh` makes loading `HADOOP_CONF_DIR` conditional on `USE_HADOOP` but `bin/zeppelin.sh` previously didn't. Should `zeppelin.sh` be made to work the same as `zeppelin-daemon.sh` or should it retain loading `HADOOP_CONF_DIR` without checking `USE_HADOOP` for backwards compatibility? * If this PR looks good, could you kindly apply the 'hacktoberfest-accepted' label? 😅 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: dev-unsubscr...@zeppelin.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
