[
https://issues.apache.org/jira/browse/HADOOP-3702?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12617633#action_12617633
]
Chris Douglas commented on HADOOP-3702:
---------------------------------------
* Instead of adding WritableUtils::asString, it might make more sense to use
the o.a.h.io.Stringifier interfaces, particularly since you create a new
JobConf (permitting a user to pass in an object would be an excellent, but not
strictly necessary, addition to Stringifier, too)
* I'm not sure I understand how the configuration deserialization works. In
Chain::getChainElementConf, a new config is created, its fields cleared and
populated from the stream, then each property defined in the deserialized conf
is (re)defined on a clone of the JobConf passed in (presumably to permit final,
etc. to be observed). Is that accurate?
* Is it true that each call to {{map}} creates a series of
{{ChainOutputCollectors}}? These look like lightweight objects, but is there a
reason the pipeline needs to be recreated each time?
* It looks like this broke the eclipse plugin:
{panel}
[exec] compile:
[exec] [echo] contrib: eclipse-plugin
[exec] [javac] Compiling 30 source files to
/zonestorage/hudson/home/hudson/hudson/jobs/Hadoop-Patch/workspace/trunk/build/contrib/eclipse-plugin/classes
[exec] [javac]
/zonestorage/hudson/home/hudson/hudson/jobs/Hadoop-Patch/workspace/trunk/src/contrib/eclipse-plugin/src/java/org/apache/hadoop/eclipse/server/HadoopServer.java:422:
write(java.io.DataOutput) in org.apache.hadoop.conf.Configuration cannot be
applied to (java.io.FileOutputStream)
[exec] [javac] this.conf.write(fos);
[exec] [javac] ^
[exec] [javac]
/zonestorage/hudson/home/hudson/hudson/jobs/Hadoop-Patch/workspace/trunk/src/contrib/eclipse-plugin/src/java/org/apache/hadoop/eclipse/servers/RunOnHadoopWizard.java:166:
write(java.io.DataOutput) in org.apache.hadoop.conf.Configuration cannot be
applied to (java.io.FileOutputStream)
[exec] [javac] conf.write(fos);
[exec] [javac] ^
{panel}
> add support for chaining Maps in a single Map and after a Reduce [M*/RM*]
> -------------------------------------------------------------------------
>
> Key: HADOOP-3702
> URL: https://issues.apache.org/jira/browse/HADOOP-3702
> Project: Hadoop Core
> Issue Type: New Feature
> Components: mapred
> Environment: all
> Reporter: Alejandro Abdelnur
> Assignee: Alejandro Abdelnur
> Priority: Minor
> Attachments: patch3702.txt, patch3702.txt, patch3702.txt,
> patch3702.txt, patch3702.txt, patch3702.txt, patch3702.txt, patch3702.txt,
> patch3702.txt
>
>
> On the same input, we usually need to run multiple Maps one after the other
> without no Reduce. We also have to run multiple Maps after the Reduce.
> If all pre-Reduce Maps are chained together and run as a single Map a
> significant amount of Disk I/O will be avoided.
> Similarly all post-Reduce Maps can be chained together and run in the Reduce
> phase after the Reduce.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
