[
https://issues.apache.org/jira/browse/HIVE-1162?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Edward Capriolo updated HIVE-1162:
----------------------------------
Description:
Hive runs atop map/reduce. Running our unit tests via map/reduce would be
faster and would make our evolution faster. Ideally we would want an ant target
in which we could supply a hadoop binary. Then we would distribute the buildir
and possibly ant via the distributed cache, finally we run all test in
parallel.
My proof of concept user bash + ssh public key.
was:
Hive run atop map/reduce, running our unit tests via map/reduce, would be
faster and would make our evolution faster. Ideally we would want an ant target
in which we could supply a hadoop binary, we would distribute the buildir and
possibly ant via the distributed cache and then run all test in parallel.
My proof of concept user bash + ssh public key.
Summary: Distribute ant test to speed process up (was: Distribute ant
test to spead process up)
> Distribute ant test to speed process up
> ---------------------------------------
>
> Key: HIVE-1162
> URL: https://issues.apache.org/jira/browse/HIVE-1162
> Project: Hadoop Hive
> Issue Type: New Feature
> Components: Build Infrastructure
> Reporter: Edward Capriolo
> Assignee: Edward Capriolo
> Priority: Minor
> Attachments: distributed_hive_test.sh
>
>
> Hive runs atop map/reduce. Running our unit tests via map/reduce would be
> faster and would make our evolution faster. Ideally we would want an ant
> target in which we could supply a hadoop binary. Then we would distribute the
> buildir and possibly ant via the distributed cache, finally we run all test
> in parallel.
> My proof of concept user bash + ssh public key.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
