[
https://issues.apache.org/jira/browse/BEAM-6114?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16913924#comment-16913924
]
Rui Wang commented on BEAM-6114:
--------------------------------
[~rahul8383]
Regarding to : // Should we throw Exception when joinType is LEFT (or) RIGHT
(or) FULL?
My perspective is, for the sake of simplicity, we can only allow triggering
once (like what CoGBK is doing). By doing so, we will allow LEFT/RIGHT/FULL
OUTER join. It is because for multiple triggering, the problem is how to refine
data. Think about outer join means it could emit <value, NULL> at the first
triggering and later it will have to emit <value, matched_value> to refine
data. We will need retractions to solve this problem.
Also it sounds nice to split javadoc of BeamJoinRel. Thanks for bringing it up.
> SQL join selection should be done in planner, not in expansion to PTransform
> ----------------------------------------------------------------------------
>
> Key: BEAM-6114
> URL: https://issues.apache.org/jira/browse/BEAM-6114
> Project: Beam
> Issue Type: Improvement
> Components: dsl-sql
> Reporter: Kenneth Knowles
> Assignee: Rahul Patwari
> Priority: Major
> Time Spent: 3.5h
> Remaining Estimate: 0h
>
> Currently Beam SQL joins all go through a single physical operator which has
> a single PTransform that does all join algorithms based on properties of its
> input PCollections as well as the relational algebra.
> A first step is to make the needed information part of the relational
> algebra, so it can choose a PTransform based on that, and the PTransforms can
> be simpler.
> Second step is to have separate (physical) relational operators for different
> join algorithms.
--
This message was sent by Atlassian Jira
(v8.3.2#803003)
