Github user ron8hu commented on a diff in the pull request:
https://github.com/apache/spark/pull/15363#discussion_r106237955
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/CostBasedJoinReorder.scala
---
@@ -51,6 +51,11 @@ case class CostBasedJoinReorder(conf: CatalystConf)
extends Rule[LogicalPlan] wi
def reorder(plan: LogicalPlan, output: AttributeSet): LogicalPlan = {
val (items, conditions) = extractInnerJoins(plan)
+ // Find the star schema joins. Currently, it returns the star join
with the largest
+ // fact table. In the future, it can return more than one star join
(e.g. F1-D1-D2
+ // and F2-D3-D4).
+ val starJoinPlans = StarSchemaDetection(conf).findStarJoins(items,
conditions.toSeq)
--- End diff --
@gatorsmile @ioana-delaney Thank you for your replies. My main point is to
identify the deficiency of the DP algorithm so that we can make improvement.
Since you are familiar with DP algorithm, can you help us identify its
deficiency/limitations so that we can improve it?
One deficiency the DP algorithm has is the explosion of the search space
when there is a large number of join relations such as >30. In
CostBasedJoinReorder, we do not optimize join order if the number of join
relations is greater than the threshold value joinReorderDPThreshold. I think
this is a place star join reorder algorithm can help. This is because it
defaults to left-deep tree which is a smaller search space. What do you think?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
