constzhou opened a new pull request #29094:
URL: https://github.com/apache/spark/pull/29094
<!--
Thanks for sending a pull request! Here are some tips for you:
1. If this is your first time, please read our contributor guidelines:
https://spark.apache.org/contributing.html
2. Ensure you have added or run the appropriate tests for your PR:
https://spark.apache.org/developer-tools.html
3. If the PR is unfinished, add '[WIP]' in your PR title, e.g.,
'[WIP][SPARK-XXXX] Your PR title ...'.
4. Be sure to keep the PR description updated to reflect all changes.
5. Please write your PR title to summarize what this PR proposes.
6. If possible, provide a concise example to reproduce the issue for a
faster review.
7. If you want to add a new configuration, please read the guideline first
for naming configurations in
'core/src/main/scala/org/apache/spark/internal/config/ConfigEntry.scala'.
-->
### What changes were proposed in this pull request?
<!--
Spark sql catalyst optimizer has a rule CollapseProject, it will try to
collapse multi projects to a single one if possible. But the collapsed project
may have more leaf expressions than the origin project. In the worst case the
number of leaf expressions will grow
exponentially, and will lead to driver oom. We could limit the max number of
leaf expression a single project can have to prevent this. The logical is that
if the project has more than 1000 leaf expressions, there is a risk to continue
collapse it. So we think this single project is very complex and we stop
collapse it. But other simple project can still collapse.
-->
### Why are the changes needed?
<!--
refer to https://issues.apache.org/jira/browse/SPARK-24983 or the unit test
added, collapse project optimizer may cause driver oom if leaf expressions grow
exponentially.
As we can see in JIRA there is a closed PR:
https://github.com/apache/spark/pull/21993 intend to fix this. There is a
comment that we should black list "case when" in collapse project, but any kind
of sql may lead to the memory problem(like the ut case), and there is a
workaround to exclude the collapse project optimizer, but I think this rule is
necessary which could improve performance. And it is hard to decide whether we
should exclude this rule ahead of time.
-->
### Does this PR introduce _any_ user-facing change?
<!--
No
-->
### How was this patch tested?
<!--
Add a unit test to check whether the query collapse to two project.
-->
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
