mirror of
https://github.com/apache/impala.git
synced 2026-01-08 03:02:48 -05:00
7906ed44ac
Implement nested-loop join in Impala with support for multiple join modes, including inner, outer, semi and anti joins. Null-aware left anti-join is not currently supported. Summary of changes: Introduced the NestedLoopJoinNode class in the FE that represents the nested loop join. Common functionality between NestedLoopJoinNode and HashJoinNode (e.g. cardinality estimation) was moved to the JoinNode class. In the BE, introduced the NestedLoopJoinNode class that implements the nested-loop join execution strategy. Change-Id: I238ec7dc0080f661847e5e1b84e30d61c3b0bb5c Reviewed-on: http://gerrit.cloudera.org:8080/652 Reviewed-by: Dimitris Tsirogiannis <dtsirogiannis@cloudera.com> Tested-by: Internal Jenkins
152 lines
5.3 KiB
Plaintext
152 lines
5.3 KiB
Plaintext
# Right outer joins with non-equi join predicates
|
|
select straight_join *
|
|
from functional.alltypestiny a right outer join functional.alltypes b
|
|
on a.id != b.id or a.int_col < b.int_col
|
|
right outer join functional.alltypesagg c
|
|
on a.smallint_col >= c.smallint_col
|
|
where a.id < 10 and c.bigint_col = 10
|
|
---- PLAN
|
|
04:NESTED LOOP JOIN [RIGHT OUTER JOIN]
|
|
| join predicates: a.smallint_col >= c.smallint_col
|
|
| predicates: a.id < 10
|
|
|
|
|
|--02:SCAN HDFS [functional.alltypesagg c]
|
|
| partitions=11/11 files=11 size=814.73KB
|
|
| predicates: c.bigint_col = 10
|
|
|
|
|
03:NESTED LOOP JOIN [RIGHT OUTER JOIN]
|
|
| join predicates: a.id != b.id OR a.int_col < b.int_col
|
|
|
|
|
|--01:SCAN HDFS [functional.alltypes b]
|
|
| partitions=24/24 files=24 size=478.45KB
|
|
|
|
|
00:SCAN HDFS [functional.alltypestiny a]
|
|
partitions=4/4 files=4 size=460B
|
|
predicates: a.id < 10
|
|
---- DISTRIBUTEDPLAN
|
|
not implemented: Error generating a valid execution plan for this query. A RIGHT OUTER JOIN type with no equi-join predicates can only be executed with a single node plan.
|
|
====
|
|
# Right semi joins with non-equi join predicates
|
|
select straight_join *
|
|
from functional.alltypestiny a right semi join functional.alltypessmall c
|
|
on a.tinyint_col > c.tinyint_col
|
|
right semi join functional.alltypesagg d
|
|
on c.tinyint_col < d.bigint_col
|
|
where d.bigint_col < 10
|
|
---- PLAN
|
|
04:NESTED LOOP JOIN [RIGHT SEMI JOIN]
|
|
| join predicates: c.tinyint_col < d.bigint_col
|
|
|
|
|
|--02:SCAN HDFS [functional.alltypesagg d]
|
|
| partitions=11/11 files=11 size=814.73KB
|
|
| predicates: d.bigint_col < 10
|
|
|
|
|
03:NESTED LOOP JOIN [RIGHT SEMI JOIN]
|
|
| join predicates: a.tinyint_col > c.tinyint_col
|
|
|
|
|
|--01:SCAN HDFS [functional.alltypessmall c]
|
|
| partitions=4/4 files=4 size=6.32KB
|
|
|
|
|
00:SCAN HDFS [functional.alltypestiny a]
|
|
partitions=4/4 files=4 size=460B
|
|
---- DISTRIBUTEDPLAN
|
|
not implemented: Error generating a valid execution plan for this query. A RIGHT SEMI JOIN type with no equi-join predicates can only be executed with a single node plan.
|
|
====
|
|
# Full outer joins with non-equi join predicates
|
|
select straight_join *
|
|
from functional.alltypestiny a full outer join functional.alltypessmall b
|
|
on a.id != b.id or a.int_col != b.int_col
|
|
full outer join functional.alltypesagg c
|
|
on a.tinyint_col > c.tinyint_col
|
|
full outer join functional.alltypes d
|
|
on c.int_col > d.int_col
|
|
where a.bigint_col != c.bigint_col and a.id < 10
|
|
---- PLAN
|
|
06:NESTED LOOP JOIN [FULL OUTER JOIN]
|
|
| join predicates: c.int_col > d.int_col
|
|
| predicates: a.bigint_col != c.bigint_col, a.id < 10
|
|
|
|
|
|--03:SCAN HDFS [functional.alltypes d]
|
|
| partitions=24/24 files=24 size=478.45KB
|
|
|
|
|
05:NESTED LOOP JOIN [FULL OUTER JOIN]
|
|
| join predicates: a.tinyint_col > c.tinyint_col
|
|
|
|
|
|--02:SCAN HDFS [functional.alltypesagg c]
|
|
| partitions=11/11 files=11 size=814.73KB
|
|
|
|
|
04:NESTED LOOP JOIN [FULL OUTER JOIN]
|
|
| join predicates: a.id != b.id OR a.int_col != b.int_col
|
|
|
|
|
|--01:SCAN HDFS [functional.alltypessmall b]
|
|
| partitions=4/4 files=4 size=6.32KB
|
|
|
|
|
00:SCAN HDFS [functional.alltypestiny a]
|
|
partitions=4/4 files=4 size=460B
|
|
predicates: a.id < 10
|
|
---- DISTRIBUTEDPLAN
|
|
not implemented: Error generating a valid execution plan for this query. A FULL OUTER JOIN type with no equi-join predicates can only be executed with a single node plan.
|
|
====
|
|
# Right anti join with non-equi join predicates
|
|
select straight_join count(*)
|
|
from functional.alltypestiny a right anti join functional.alltypessmall b
|
|
on a.id < b.id
|
|
where b.int_col = 5
|
|
---- PLAN
|
|
03:AGGREGATE [FINALIZE]
|
|
| output: count(*)
|
|
|
|
|
02:NESTED LOOP JOIN [RIGHT ANTI JOIN]
|
|
| join predicates: a.id < b.id
|
|
|
|
|
|--01:SCAN HDFS [functional.alltypessmall b]
|
|
| partitions=4/4 files=4 size=6.32KB
|
|
| predicates: b.int_col = 5
|
|
|
|
|
00:SCAN HDFS [functional.alltypestiny a]
|
|
partitions=4/4 files=4 size=460B
|
|
---- DISTRIBUTEDPLAN
|
|
not implemented: Error generating a valid execution plan for this query. A RIGHT ANTI JOIN type with no equi-join predicates can only be executed with a single node plan.
|
|
====
|
|
# Inner and right joins with non-equi join predicates
|
|
select straight_join count(*)
|
|
from functional.alltypestiny a inner join functional.alltypessmall b on a.id < b.id
|
|
right outer join functional.alltypesagg c on a.int_col != c.int_col
|
|
right semi join functional.alltypes d on c.tinyint_col < d.tinyint_col
|
|
right anti join functional.alltypesnopart e on d.tinyint_col > e.tinyint_col
|
|
where e.id < 10
|
|
---- PLAN
|
|
09:AGGREGATE [FINALIZE]
|
|
| output: count(*)
|
|
|
|
|
08:NESTED LOOP JOIN [RIGHT ANTI JOIN]
|
|
| join predicates: d.tinyint_col > e.tinyint_col
|
|
|
|
|
|--04:SCAN HDFS [functional.alltypesnopart e]
|
|
| partitions=1/1 files=0 size=0B
|
|
| predicates: e.id < 10
|
|
|
|
|
07:NESTED LOOP JOIN [RIGHT SEMI JOIN]
|
|
| join predicates: c.tinyint_col < d.tinyint_col
|
|
|
|
|
|--03:SCAN HDFS [functional.alltypes d]
|
|
| partitions=24/24 files=24 size=478.45KB
|
|
|
|
|
06:NESTED LOOP JOIN [RIGHT OUTER JOIN]
|
|
| join predicates: a.int_col != c.int_col
|
|
|
|
|
|--02:SCAN HDFS [functional.alltypesagg c]
|
|
| partitions=11/11 files=11 size=814.73KB
|
|
|
|
|
05:NESTED LOOP JOIN [INNER JOIN]
|
|
| predicates: a.id < b.id
|
|
|
|
|
|--01:SCAN HDFS [functional.alltypessmall b]
|
|
| partitions=4/4 files=4 size=6.32KB
|
|
|
|
|
00:SCAN HDFS [functional.alltypestiny a]
|
|
partitions=4/4 files=4 size=460B
|
|
---- DISTRIBUTEDPLAN
|
|
not implemented: Error generating a valid execution plan for this query. A RIGHT ANTI JOIN type with no equi-join predicates can only be executed with a single node plan.
|
|
====
|