[
https://issues.apache.org/jira/browse/PIG-157?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12594925#action_12594925
]
Pi Song commented on PIG-157:
-----------------------------
Good to see people start writing ideas to Wiki!!
1) First concern is that using Hadoop Local will tie us to Hadoop too much.
There was an initiative quite a while ago to start looking at different
backends other than Hadoop (e.g. we might be running a backend like [EMAIL
PROTECTED] Who knows?).
However, this whole thing seems to have been built for solely Hadoop anyway.
Not sure about the current direction.
2) Have you tried to measure LocalHadoop startup time compared to the local
engine? If the LocalHadoop takes much more time to startup, we might suffer
when processing nested queries.
> Add types and rework execution pipeline
> ---------------------------------------
>
> Key: PIG-157
> URL: https://issues.apache.org/jira/browse/PIG-157
> Project: Pig
> Issue Type: New Feature
> Components: impl
> Reporter: Alan Gates
> Assignee: Alan Gates
> Attachments: Core.patch.zip, exceptions.patch, incr1.zip
>
>
> This is the tracking bug for the work to add types to pig and rework the
> execution pipeline. Individual components of this work are covered in
> subtasks.
> Functional and design specs for this work are:
> http://wiki.apache.org/pig/PigTypesFunctionalSpec
> http://wiki.apache.org/pig/PigTypesDesign
> http://wiki.apache.org/pig/PigExecutionModel
> This work is being done on the branch types, since it is large and
> disruptive, and we want to be able to do incremental checkins without causing
> issues for the trunk.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
