-
Notifications
You must be signed in to change notification settings - Fork 3.7k
[fix](nereids)add a project node above sort node to eliminate unused order by keys #17913
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
593babf to
839394b
Compare
|
run buildall |
|
TeamCity pipeline, clickbench performance test result: |
839394b to
7be2464
Compare
|
run buildall |
7be2464 to
64bfc78
Compare
|
run buildall |
|
the column pruning problem will be complete fixed by #17579 |
64bfc78 to
1abe6b2
Compare
|
run buildall |
|
PR approved by at least one committer and no changes requested. |
|
PR approved by anyone and no changes requested. |
… order by keys (apache#17913) if the order by keys are not simple slot in sort node, the order by exprs have to been added to sort node's output tuple. In that case, we need add a project node above sort node to eliminate the unused order by exprs. for example: ```sql WITH t0 AS (SELECT DATE_FORMAT(date, '%Y%m%d') AS date FROM cir_1756_t1 ), t3 AS (SELECT date_format(date, '%Y%m%d') AS `date` FROM `cir_1756_t2` GROUP BY date_format(date, '%Y%m%d') **ORDER BY date_format(date, '%Y%m%d')** ) SELECT t0.date FROM t0 LEFT JOIN t3 ON t0.date = t3.date; ``` before: ``` +--------------------------------------------------------------------------------------------------------------------------------------------------+ | Explain String | +--------------------------------------------------------------------------------------------------------------------------------------------------+ | LogicalProject[159] ( distinct=false, projects=[date#1], excepts=[], canEliminate=true ) | | +--LogicalJoin[158] ( type=LEFT_OUTER_JOIN, markJoinSlotReference=Optional.empty, hashJoinConjuncts=[(date#1 = date#3)], otherJoinConjuncts=[] ) | | |--LogicalProject[151] ( distinct=false, projects=[date_format(date#0, '%Y%m%d') AS `date`apache#1], excepts=[], canEliminate=true ) | | | +--LogicalOlapScan ( qualified=default_cluster:bugfix.cir_1756_t1, indexName=cir_1756_t1, selectedIndexId=412339, preAgg=ON ) | | +--LogicalSort[157] ( orderKeys=[date_format(cast(date#3 as DATETIME), '%Y%m%d') asc null first] ) | | +--LogicalAggregate[156] ( groupByExpr=[date#3], outputExpr=[date#3], hasRepeat=false ) | | +--LogicalProject[155] ( distinct=false, projects=[date_format(date#2, '%Y%m%d') AS `date`apache#3], excepts=[], canEliminate=true ) | | +--LogicalOlapScan ( qualified=default_cluster:bugfix.cir_1756_t2, indexName=cir_1756_t2, selectedIndexId=412352, preAgg=ON ) | +--------------------------------------------------------------------------------------------------------------------------------------------------+ ``` after: ``` +--------------------------------------------------------------------------------------------------------------------------------------------------+ | Explain String | +--------------------------------------------------------------------------------------------------------------------------------------------------+ | LogicalProject[171] ( distinct=false, projects=[date#2], excepts=[], canEliminate=true ) | | +--LogicalJoin[170] ( type=LEFT_OUTER_JOIN, markJoinSlotReference=Optional.empty, hashJoinConjuncts=[(date#2 = date#4)], otherJoinConjuncts=[] ) | | |--LogicalProject[162] ( distinct=false, projects=[date_format(date#0, '%Y%m%d') AS `date`apache#2], excepts=[], canEliminate=true ) | | | +--LogicalOlapScan ( qualified=default_cluster:bugfix.cir_1756_t1, indexName=cir_1756_t1, selectedIndexId=1049812, preAgg=ON ) | | +--LogicalProject[169] ( distinct=false, projects=[date#4], excepts=[], canEliminate=false ) | | +--LogicalSort[168] ( orderKeys=[date_format(cast(date#4 as DATETIME), '%Y%m%d') asc null first] ) | | +--LogicalAggregate[167] ( groupByExpr=[date#4], outputExpr=[date#4], hasRepeat=false ) | | +--LogicalProject[166] ( distinct=false, projects=[date_format(date#3, '%Y%m%d') AS `date`apache#4], excepts=[], canEliminate=true ) | | +--LogicalOlapScan ( qualified=default_cluster:bugfix.cir_1756_t2, indexName=cir_1756_t2, selectedIndexId=1049825, preAgg=ON ) | +--------------------------------------------------------------------------------------------------------------------------------------------------+ ```
Proposed changes
Issue Number: close #xxx
Problem summary
if the order by keys are not simple slot in sort node, the order by exprs have to been added to sort node's output tuple. In that case, we need add a project node above sort node to eliminate the unused order by exprs. for example:
before:
after:
Checklist(Required)
Further comments
If this is a relatively large or complex change, kick off the discussion at dev@doris.apache.org by explaining why you chose the solution you did and what alternatives you considered, etc...