Github user srowen commented on the pull request:
https://github.com/apache/spark/pull/629#issuecomment-42644509
`json4s` is a direct dependency of Spark core. It's not related to the
changes here. It sounds like something is running but being pointed to an
artifact that does not contain dependencies, and they are not otherwise on the
classpath.
I am not sure how you're modifying/running this, but I _suspect_ you
continue to collide with existing 0.9.0 binaries and environment variables on
your cluster. Building matching binaries is the easy part. It's deploying it
manually that may be difficult. It should just be a matter of scraping out and
replacing the stuff you find in `/opt/cloudera/.../spark`, but you may have to
be sure to replace many things, not just one assembly jar. You'd have to
restart the services too. And this has to happen on all workers.
Longer-term of course an updated Spark will be deployed with CDH updates
anyway.
If you just want to move off Hive, there's another answer for that, but
that's a different question.
---
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.
---
