mirror of
https://github.com/apache/impala.git
synced 2026-01-06 15:01:43 -05:00
9f61397fc4
The plan-root fragment instance that runs on the coordinator should be handled like all others: started via RPC and run asynchronously. Without this, the fragment requires special-case code throughout the coordinator, and does not show up in system metrics etc. This patch adds a new sink type, PlanRootSink, to the root fragment instance so that the coordinator can pull row batches that are pushed by the root instance. The coordinator signals completion to the fragment instance via closing the consumer side of the sink, whereupon the instance is free to complete. Since the root instance now runs asynchronously wrt to the coordinator, we add several coordination methods to allow the coordinator to wait for a point in the instance's execution to be hit - e.g. to wait until the instance has been opened. Done in this patch: * Add PlanRootSink * Add coordination to PFE to allow coordinator to observe lifecycle * Make FragmentMgr a singleton * Removed dead code from Coordinator::Wait() and elsewhere. * Moved result output exprs out of QES and into PlanRootSink. * Remove special-case limit-based teardown of coordinator fragment, and supporting functions in PlanFragmentExecutor. * Simplified lifecycle of PlanFragmentExecutor by separating Open() into Open() and Exec(), the latter of which drives the sink by reading rows from the plan tree. * Add child profile to PlanFragmentExecutor to measure time spent in each lifecycle phase. * Removed dependency between InitExecProfiles() and starting root fragment. * Removed mostly dead-code handling of LIMIT 0 queries. * Ensured that SET returns a result set in all cases. * Fix test_get_log() HS2 test. Errors are only guaranteed to be visible after fetch calls return EOS, but test was assuming this would happen after first fetch. Change-Id: Ibb0064ec2f085fa3a5598ea80894fb489a01e4df Reviewed-on: http://gerrit.cloudera.org:8080/4402 Tested-by: Internal Jenkins Reviewed-by: Henry Robinson <henry@cloudera.com>
217 lines
7.1 KiB
Plaintext
217 lines
7.1 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
|
|
PLAN-ROOT SINK
|
|
|
|
|
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
|
|
PLAN-ROOT SINK
|
|
|
|
|
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
|
|
PLAN-ROOT SINK
|
|
|
|
|
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
|
|
PLAN-ROOT SINK
|
|
|
|
|
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
|
|
PLAN-ROOT SINK
|
|
|
|
|
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.
|
|
====
|
|
# Right semi and outer joins are inverted to make them executable.
|
|
# Same query as above but without the straight join hint.
|
|
select 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
|
|
---- DISTRIBUTEDPLAN
|
|
PLAN-ROOT SINK
|
|
|
|
|
15:AGGREGATE [FINALIZE]
|
|
| output: count:merge(*)
|
|
|
|
|
14:EXCHANGE [UNPARTITIONED]
|
|
|
|
|
09:AGGREGATE
|
|
| output: count(*)
|
|
|
|
|
08:NESTED LOOP JOIN [LEFT ANTI JOIN, BROADCAST]
|
|
| join predicates: d.tinyint_col > e.tinyint_col
|
|
|
|
|
|--13:EXCHANGE [BROADCAST]
|
|
| |
|
|
| 07:NESTED LOOP JOIN [LEFT SEMI JOIN, BROADCAST]
|
|
| | join predicates: c.tinyint_col < d.tinyint_col
|
|
| |
|
|
| |--12:EXCHANGE [BROADCAST]
|
|
| | |
|
|
| | 06:NESTED LOOP JOIN [LEFT OUTER JOIN, BROADCAST]
|
|
| | | join predicates: a.int_col != c.int_col
|
|
| | |
|
|
| | |--11:EXCHANGE [BROADCAST]
|
|
| | | |
|
|
| | | 05:NESTED LOOP JOIN [INNER JOIN, BROADCAST]
|
|
| | | | predicates: a.id < b.id
|
|
| | | |
|
|
| | | |--10:EXCHANGE [BROADCAST]
|
|
| | | | |
|
|
| | | | 00:SCAN HDFS [functional.alltypestiny a]
|
|
| | | | partitions=4/4 files=4 size=460B
|
|
| | | |
|
|
| | | 01:SCAN HDFS [functional.alltypessmall b]
|
|
| | | partitions=4/4 files=4 size=6.32KB
|
|
| | |
|
|
| | 02:SCAN HDFS [functional.alltypesagg c]
|
|
| | partitions=11/11 files=11 size=814.73KB
|
|
| |
|
|
| 03:SCAN HDFS [functional.alltypes d]
|
|
| partitions=24/24 files=24 size=478.45KB
|
|
|
|
|
04:SCAN HDFS [functional.alltypesnopart e]
|
|
partitions=1/1 files=0 size=0B
|
|
predicates: e.id < 10
|
|
====
|