[DRILL-2639] Planner bug - RelOptPlanner.CannotPlanException

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 0.9.0
• Fix Version/s: 0.9.0
• Component/s: Query Planning & Optimization

Environment

9d92b8e319f2d46e8659d903d355450e15946533

DRILL-2580: Exit early from HashJoinBatch if build side is empty

26.03.2015 @ 16:13:53 EDT

Unknown

26.03.2015 @ 16:53:21 EDT

• Assignee: Sean Hsuan-Yi Chu
• Reporter: Khurram Faraaz
• Created: 31/Mar/15 20:57
• Updated: 10/Dec/15 18:49
• Resolved: 10/Apr/15 23:23

Description

Reporting this as a separate JIRA as this issue related to a bug in the planner. Performing aggregate on the output returned by Union All results in CannotPlanException. Note that the two inputs to Union All are casted to integer and hence the inputs from both legs are of the same datatype.

0: jdbc:drill:> select count(c1) from (select cast(columns[0] as int) c1 from `testWindow.csv`) union all (select cast(columns[0] as int) c2 from `testWindow.csv`);
Query failed: RelOptPlanner.CannotPlanException: Node [rel#59393:Subset#4.LOGICAL.ANY([]).[]] could not be implemented; planner state:

Root: rel#59393:Subset#4.LOGICAL.ANY([]).[]
Original rel:
AbstractConverter(subset=[rel#59393:Subset#4.LOGICAL.ANY([]).[]], convention=[LOGICAL], DrillDistributionTraitDef=[ANY([])], sort=[[]]): rowcount = 1.7976931348623157E308, cumulative cost = {inf}, id = 59394
  UnionRel(subset=[rel#59392:Subset#4.NONE.ANY([]).[]], all=[true]): rowcount = 1.7976931348623157E308, cumulative cost = {1.7976931348623157E308 rows, 1.7976931348623157E308 cpu, 0.0 io, 0.0 network, 0.0 memory}, id = 59391
    AggregateRel(subset=[rel#59388:Subset#2.NONE.ANY([]).[]], group=[{}], EXPR$0=[COUNT($0)]): rowcount = 1.7976931348623158E307, cumulative cost = {1.7976931348623158E307 rows, 0.0 cpu, 0.0 io, 0.0 network, 0.0 memory}, id = 59387
      ProjectRel(subset=[rel#59386:Subset#1.NONE.ANY([]).[]], c1=[CAST(ITEM($1, 0)):INTEGER]): rowcount = 100.0, cumulative cost = {100.0 rows, 100.0 cpu, 0.0 io, 0.0 network, 0.0 memory}, id = 59385
        EnumerableTableAccessRel(subset=[rel#59384:Subset#0.ENUMERABLE.ANY([]).[]], table=[[dfs, tmp, testWindow.csv]]): rowcount = 100.0, cumulative cost = {100.0 rows, 101.0 cpu, 0.0 io, 0.0 network, 0.0 memory}, id = 59368
    ProjectRel(subset=[rel#59390:Subset#3.NONE.ANY([]).[]], c2=[CAST(ITEM($1, 0)):INTEGER]): rowcount = 100.0, cumulative cost = {100.0 rows, 100.0 cpu, 0.0 io, 0.0 network, 0.0 memory}, id = 59389
      EnumerableTableAccessRel(subset=[rel#59384:Subset#0.ENUMERABLE.ANY([]).[]], table=[[dfs, tmp, testWindow.csv]]): rowcount = 100.0, cumulative cost = {100.0 rows, 101.0 cpu, 0.0 io, 0.0 network, 0.0 memory}, id = 59368

Sets:
Set#0, type: (DrillRecordRow[*, columns])
	rel#59384:Subset#0.ENUMERABLE.ANY([]).[], best=rel#59368, importance=0.6561
		rel#59368:EnumerableTableAccessRel.ENUMERABLE.ANY([]).[](table=[dfs, tmp, testWindow.csv]), rowcount=100.0, cumulative cost={100.0 rows, 101.0 cpu, 0.0 io, 0.0 network, 0.0 memory}
		rel#59408:AbstractConverter.ENUMERABLE.ANY([]).[](child=rel#59407:Subset#0.LOGICAL.ANY([]).[],convention=ENUMERABLE,DrillDistributionTraitDef=ANY([]),sort=[]), rowcount=1.0, cumulative cost={inf}
	rel#59407:Subset#0.LOGICAL.ANY([]).[], best=rel#59415, importance=0.5904900000000001
		rel#59409:AbstractConverter.LOGICAL.ANY([]).[](child=rel#59384:Subset#0.ENUMERABLE.ANY([]).[],convention=LOGICAL,DrillDistributionTraitDef=ANY([]),sort=[]), rowcount=100.0, cumulative cost={inf}
		rel#59415:DrillScanRel.LOGICAL.ANY([]).[](table=[dfs, tmp, testWindow.csv],groupscan=EasyGroupScan [selectionRoot=/tmp/testWindow.csv, numFiles=1, columns=[`*`], files=[maprfs:/tmp/testWindow.csv]]), rowcount=1.0, cumulative cost={1.0 rows, 10000.0 cpu, 0.0 io, 0.0 network, 0.0 memory}
Set#1, type: RecordType(INTEGER c1)
	rel#59386:Subset#1.NONE.ANY([]).[], best=null, importance=0.7290000000000001
		rel#59385:ProjectRel.NONE.ANY([]).[](child=rel#59384:Subset#0.ENUMERABLE.ANY([]).[],c1=CAST(ITEM($1, 0)):INTEGER), rowcount=100.0, cumulative cost={inf}
	rel#59404:Subset#1.LOGICAL.ANY([]).[], best=rel#59413, importance=0.36450000000000005
		rel#59405:AbstractConverter.LOGICAL.ANY([]).[](child=rel#59386:Subset#1.NONE.ANY([]).[],convention=LOGICAL,DrillDistributionTraitDef=ANY([]),sort=[]), rowcount=1.7976931348623157E308, cumulative cost={inf}
		rel#59413:DrillProjectRel.LOGICAL.ANY([]).[](child=rel#59402:Subset#5.LOGICAL.ANY([]).[],c1=CAST(ITEM($0, 0)):INTEGER), rowcount=1.0, cumulative cost={2.0 rows, 5.0 cpu, 0.0 io, 0.0 network, 0.0 memory}
		rel#59414:DrillProjectRel.LOGICAL.ANY([]).[](child=rel#59407:Subset#0.LOGICAL.ANY([]).[],c1=CAST(ITEM($1, 0)):INTEGER), rowcount=1.0, cumulative cost={2.0 rows, 10004.0 cpu, 0.0 io, 0.0 network, 0.0 memory}
Set#2, type: RecordType(BIGINT EXPR$0)
	rel#59388:Subset#2.NONE.ANY([]).[], best=null, importance=0.81
		rel#59387:AggregateRel.NONE.ANY([]).[](child=rel#59386:Subset#1.NONE.ANY([]).[],group={},EXPR$0=COUNT($0)), rowcount=1.7976931348623158E307, cumulative cost={inf}
	rel#59395:Subset#2.LOGICAL.ANY([]).[], best=rel#59406, importance=0.405
		rel#59396:AbstractConverter.LOGICAL.ANY([]).[](child=rel#59388:Subset#2.NONE.ANY([]).[],convention=LOGICAL,DrillDistributionTraitDef=ANY([]),sort=[]), rowcount=1.7976931348623157E308, cumulative cost={inf}
		rel#59406:DrillAggregateRel.LOGICAL.ANY([]).[](child=rel#59404:Subset#1.LOGICAL.ANY([]).[],group={},EXPR$0=COUNT($0)), rowcount=1.0, cumulative cost={3.0 rows, 6.0 cpu, 0.0 io, 0.0 network, 0.0 memory}
Set#3, type: RecordType(INTEGER c2)
	rel#59390:Subset#3.NONE.ANY([]).[], best=null, importance=0.81
		rel#59389:ProjectRel.NONE.ANY([]).[](child=rel#59384:Subset#0.ENUMERABLE.ANY([]).[],c2=CAST(ITEM($1, 0)):INTEGER), rowcount=100.0, cumulative cost={inf}
	rel#59397:Subset#3.LOGICAL.ANY([]).[], best=rel#59403, importance=0.405
		rel#59398:AbstractConverter.LOGICAL.ANY([]).[](child=rel#59390:Subset#3.NONE.ANY([]).[],convention=LOGICAL,DrillDistributionTraitDef=ANY([]),sort=[]), rowcount=1.7976931348623157E308, cumulative cost={inf}
		rel#59403:DrillProjectRel.LOGICAL.ANY([]).[](child=rel#59402:Subset#5.LOGICAL.ANY([]).[],c2=CAST(ITEM($0, 0)):INTEGER), rowcount=1.0, cumulative cost={2.0 rows, 5.0 cpu, 0.0 io, 0.0 network, 0.0 memory}
		rel#59410:DrillProjectRel.LOGICAL.ANY([]).[](child=rel#59407:Subset#0.LOGICAL.ANY([]).[],c2=CAST(ITEM($1, 0)):INTEGER), rowcount=1.0, cumulative cost={2.0 rows, 10004.0 cpu, 0.0 io, 0.0 network, 0.0 memory}
Set#4, type: RecordType(BIGINT EXPR$0)
	rel#59392:Subset#4.NONE.ANY([]).[], best=null, importance=0.9
		rel#59391:UnionRel.NONE.ANY([]).[](input#0=rel#59388:Subset#2.NONE.ANY([]).[],input#1=rel#59390:Subset#3.NONE.ANY([]).[],all=true), rowcount=1.7976931348623157E308, cumulative cost={inf}
	rel#59393:Subset#4.LOGICAL.ANY([]).[], best=null, importance=1.0
		rel#59394:AbstractConverter.LOGICAL.ANY([]).[](child=rel#59392:Subset#4.NONE.ANY([]).[],convention=LOGICAL,DrillDistributionTraitDef=ANY([]),sort=[]), rowcount=1.7976931348623157E308, cumulative cost={inf}
Set#5, type: RecordType(ANY columns)
	rel#59402:Subset#5.LOGICAL.ANY([]).[], best=rel#59400, importance=0.12728571428571428
		rel#59400:DrillScanRel.LOGICAL.ANY([]).[](table=[dfs, tmp, testWindow.csv],groupscan=EasyGroupScan [selectionRoot=/tmp/testWindow.csv, numFiles=1, columns=[`columns`[0]], files=[maprfs:/tmp/testWindow.csv]]), rowcount=1.0, cumulative cost={1.0 rows, 1.0 cpu, 0.0 io, 0.0 network, 0.0 memory}

Error: exception while executing query: Failure while executing query. (state=,code=0)

Stack trace from drillbit.log

Set#5, type: RecordType(ANY columns)
        rel#59402:Subset#5.LOGICAL.ANY([]).[], best=rel#59400, importance=0.12728571428571428
                rel#59400:DrillScanRel.LOGICAL.ANY([]).[](table=[dfs, tmp, testWindow.csv],groupscan=EasyGroupScan [selectionRoot=/tmp/testWindow.csv, numFiles=1, columns=[`columns`[0]], files=[maprfs:/tmp/testWindow.csv]]), rowcount=1.0, cumulative cost={1.0 rows, 1.0 cpu, 0.0 io, 0.0 network, 0.0 memory}


        at org.eigenbase.relopt.volcano.RelSubset$CheapestPlanReplacer.visit(RelSubset.java:445) ~[optiq-core-0.9-drill-r20.jar:na]
        at org.eigenbase.relopt.volcano.RelSubset.buildCheapestPlan(RelSubset.java:287) ~[optiq-core-0.9-drill-r20.jar:na]
        at org.eigenbase.relopt.volcano.VolcanoPlanner.findBestExp(VolcanoPlanner.java:677) ~[optiq-core-0.9-drill-r20.jar:na]
        at net.hydromatic.optiq.tools.Programs$RuleSetProgram.run(Programs.java:165) ~[optiq-core-0.9-drill-r20.jar:na]
        at net.hydromatic.optiq.prepare.PlannerImpl.transform(PlannerImpl.java:275) ~[optiq-core-0.9-drill-r20.jar:na]
        at org.apache.drill.exec.planner.sql.handlers.DefaultSqlHandler.convertToDrel(DefaultSqlHandler.java:206) ~[drill-java-exec-0.9.0-SNAPSHOT-rebuffed.jar:0.9.0-SNAPSHOT]
        at org.apache.drill.exec.planner.sql.handlers.DefaultSqlHandler.getPlan(DefaultSqlHandler.java:138) ~[drill-java-exec-0.9.0-SNAPSHOT-rebuffed.jar:0.9.0-SNAPSHOT]
        at org.apache.drill.exec.planner.sql.DrillSqlWorker.getPlan(DrillSqlWorker.java:145) ~[drill-java-exec-0.9.0-SNAPSHOT-rebuffed.jar:0.9.0-SNAPSHOT]
        at org.apache.drill.exec.work.foreman.Foreman.runSQL(Foreman.java:773) [drill-java-exec-0.9.0-SNAPSHOT-rebuffed.jar:0.9.0-SNAPSHOT]
        at org.apache.drill.exec.work.foreman.Foreman.run(Foreman.java:204) [drill-java-exec-0.9.0-SNAPSHOT-rebuffed.jar:0.9.0-SNAPSHOT]
        ... 3 common frames omitted