[
https://issues.apache.org/jira/browse/FLINK-34702?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17883560#comment-17883560
]
Sergey Nuyanzin edited comment on FLINK-34702 at 9/22/24 12:08 AM:
-------------------------------------------------------------------
I had a look here a bit.
I noticed that there is a number of similar cases where this
{quote}StreamPhysicalDeduplicate doesn't support consuming update changes which
is produced by node{quote} is happening and node could be also a join node like
left/right/full outer join, semijoin, antijoin. Look at the end to see examples.
Since there are some difficulties mentioned above with also 3 aforementioned
solutions, how about going from a bit different direction?
What if we in {{StreamPhysicalRankRule#convert}} do not convert to deduplicate
{{RelNode}}'s if they in input contains {{GroupAggregate}}'s,
{{LeftOuterJoin}}, {{RightOuterJoin}}, {{FullOuterJoin}}, {{SemiJoin}},
{{AntiJoin}} ?
I submitted a draft PR with this proposal (first to make sure that it passes
existing ci tests)
More examples of failing queries
FullOuterJoin (Left/RightOuterJoin are similar)
{code:sql}
SELECT *
FROM (
SELECT *,
ROW_NUMBER() OVER (PARTITION BY a ORDER BY PROCTIME() ASC) as rowNum
FROM (SELECT m.a, m.b FROM MyTable m FULL OUTER JOIN MyTable2 m2 ON m.a =
m2.a)
)
WHERE rowNum = 1
{code}
SemiJoin
{code:sql}
SELECT *
FROM (
SELECT *,
ROW_NUMBER() OVER (PARTITION BY a ORDER BY PROCTIME() ASC) as rowNum
FROM (SELECT a, b FROM MyTable
WHERE EXISTS (
SELECT a, b FROM MyTable2))
)
WHERE rowNum = 1
{code}
AntiJoin
{code:sql}
SELECT *
FROM (
SELECT *,
ROW_NUMBER() OVER (PARTITION BY a ORDER BY PROCTIME() ASC) as rowNum
FROM (SELECT a, b FROM MyTable
WHERE NOT EXISTS (
SELECT a, b FROM MyTable2))
)
WHERE rowNum = 1
{code}
was (Author: sergey nuyanzin):
I had a look here a bit.
I noticed that there is a number of similar cases where this
{quote}StreamPhysicalDeduplicate doesn't support consuming update changes which
is produced by node{quote} is happening and node could be also a join node like
left/right/full outer join, semijoin, antijoin. Look at the end to see examples.
Since there are some difficulties mentioned above with also 3 aforementioned
solutions, how about going from a bit different direction?
What if we in {{StreamPhysicalRankRule#convert}} do not convert to deduplicate
{{RelNode}}'s if they in input contains {{GroupAggregate}}'s,
{{LeftOuterJoin}}, {{RightOuterJoin}}, {{FullOuterJoin}}, {{SemiJoin}},
{{AntiJoin}} ?
I submitted a draft PR with this proposal (first to make sure that it passes
existing ci tests)
FullOuterJoin (Left/RightOuterJoin are similar)
{code:sql}
SELECT *
FROM (
SELECT *,
ROW_NUMBER() OVER (PARTITION BY a ORDER BY PROCTIME() ASC) as rowNum
FROM (SELECT m.a, m.b FROM MyTable m FULL OUTER JOIN MyTable2 m2 ON m.a =
m2.a)
)
WHERE rowNum = 1
{code}
SemiJoin
{code:sql}
SELECT *
FROM (
SELECT *,
ROW_NUMBER() OVER (PARTITION BY a ORDER BY PROCTIME() ASC) as rowNum
FROM (SELECT a, b FROM MyTable
WHERE EXISTS (
SELECT a, b FROM MyTable2))
)
WHERE rowNum = 1
{code}
AntiJoin
{code:sql}
SELECT *
FROM (
SELECT *,
ROW_NUMBER() OVER (PARTITION BY a ORDER BY PROCTIME() ASC) as rowNum
FROM (SELECT a, b FROM MyTable
WHERE NOT EXISTS (
SELECT a, b FROM MyTable2))
)
WHERE rowNum = 1
{code}
> Rank should not convert to StreamExecDuplicate when the input is not insert
> only
> --------------------------------------------------------------------------------
>
> Key: FLINK-34702
> URL: https://issues.apache.org/jira/browse/FLINK-34702
> Project: Flink
> Issue Type: Bug
> Components: Table SQL / Planner
> Affects Versions: 1.20.0
> Reporter: Jacky Lau
> Priority: Major
> Labels: pull-request-available
> Fix For: 2.0.0
>
>
> {code:java}
> @Test
> def testSimpleFirstRowOnBuiltinProctime1(): Unit = {
> val sqlQuery =
> """
> |SELECT *
> |FROM (
> | SELECT *,
> | ROW_NUMBER() OVER (PARTITION BY a ORDER BY PROCTIME() ASC) as
> rowNum
> | FROM (select a, count(b) as b from MyTable group by a)
> |)
> |WHERE rowNum = 1
> """.stripMargin
> util.verifyExecPlan(sqlQuery)
> } {code}
> Exception:
> org.apache.flink.table.api.TableException: StreamPhysicalDeduplicate doesn't
> support consuming update changes which is produced by node
> GroupAggregate(groupBy=[a], select=[a, COUNT(b) AS b])
> because the StreamPhysicalDeduplicate can not consuming update changes now
> while StreamExecRank can.
> so we should not convert the FlinkLogicalRank to StreamPhysicalDeduplicate in
> this case. and we can defer whether input contains update change in the
> "optimize the physical plan" phase.
> so we can add an option to solve it. and when the StreamPhysicalDeduplicate
> can support consuming update changes , we can deprecate it
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
