Github user srowen commented on the pull request:
https://github.com/apache/spark/pull/9759#issuecomment-220270486
Hm, I know that Hadoop libs use Jackson 2.4. Including and using a later
version can mostly work, especially if Spark code doesn't use APIs from later
versions. I think we concluded 2.5 was fine and updated to it for 2.x; I don't
know about 2.7. I know I have had problems using even Jackson 2.5 in the past
with Spark apps.
We should probably forge ahead and see what builds like the Hadoop 2.2
builds do. If there are problems, we may have to fix-forward by shading Jackson
rather than get stuck on Jackson 2.5.
It would be great to check whether _all_ the Jackson dependencies are being
managed up to 2.7.3. For example if a transitive dependency depends on some
Jackson module that we don't manage the version of, it might be coming in still
with an older version. `mvn dependency:tree` and grepping for "jackson" and
then for versions that aren't 2.7.3 would figure that out. Ideally try `sbt
dependency-graph` too.
I'm surprised that the paranamer dependency went down in version. The
exercise above may reveal why -- what is bringing it in at 2.3 now?
---
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.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
