Github user concretevitamin commented on the pull request:
https://github.com/apache/spark/pull/956#issuecomment-45279750
I removed a unit test introduced in this PR which was failing in the most
recent build. That test was ensuring SQLConf is always a subset of HiveConf.
However, the keys `datanucleus.cache.collections.lazy`,
`datanucleus.cache.collections`,
`hive.metastore.partition.name.whitelist.pattern` are executed by Hive in
various queries in the test suite, and hence are not picked up by SQLConf but
reside in HiveConf.
One proposal is to just remove the contract that SQLConf is a subset of
HiveConf, and instead merely enforce any SET by `sql()/hql()` and
`sqlConf.set()` will be picked by HiveConf, in the case where a HiveContext is
in use. This sounds reasonable to me as we don't have an immediate use case
where a subset relation is depended upon on (or do we?).
---
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.
---
