[
https://issues.apache.org/jira/browse/CRUNCH-388?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Josh Wills updated CRUNCH-388:
------------------------------
Attachment: CRUNCH-388.patch
The fixes:
1) Added getNumReduceTasks() support to the fake TaskInputOutputContext we use
in the MemoryCollection.
2) Be more careful about setting the JARs on the JavaSparkContext and logging
any errors we encounter when we're running.
> Fix issues running Spark/memory impls against Oryx
> --------------------------------------------------
>
> Key: CRUNCH-388
> URL: https://issues.apache.org/jira/browse/CRUNCH-388
> Project: Crunch
> Issue Type: Bug
> Affects Versions: 0.9.0
> Reporter: Josh Wills
> Fix For: 0.10.0
>
> Attachments: CRUNCH-388.patch
>
>
> I found a couple of small issues with the in-memory implementation and the
> Spark implementation when I was testing them out against Oryx, a machine
> learning project I work on that uses Crunch's MR pipeline implementation.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
