[jira] [Created] (FLINK-15555) Delete TABLE_OPTIMIZER_REUSE_SOURCE_ENABLED option for subplaner reuse

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[jira] [Created] (FLINK-15555) Delete TABLE_OPTIMIZER_REUSE_SOURCE_ENABLED option for subplaner reuse

Shang Yuanchun (Jira)
hailong wang created FLINK-15555:
------------------------------------

             Summary: Delete TABLE_OPTIMIZER_REUSE_SOURCE_ENABLED  option for subplaner reuse
                 Key: FLINK-15555
                 URL: https://issues.apache.org/jira/browse/FLINK-15555
             Project: Flink
          Issue Type: Improvement
          Components: Table SQL / Runtime
    Affects Versions: 1.10.0
            Reporter: hailong wang
             Fix For: 1.11.0


Blink planner supports subplan reuse. If TABLE_OPTIMIZER_REUSE_SUB_PLAN_ENABLED is true, the optimizer will try to find out duplicated sub-plans and reuse them. And  if TABLE_OPTIMIZER_REUSE_SOURCE_ENABLED is true, the optimizer will try to find out duplicated table sources and reuse them.

The option of TABLE_OPTIMIZER_REUSE_SOURCE_ENABLED used to defined whether TableSourceScan should be reused.

But if  the parent's relNode of TableSourceScan can be reused, it will be also reused even if TABLE_OPTIMIZER_REUSE_SOURCE_ENABLED is false, just like follow sql:
{code:java}
WITH t AS (SELECT a, b, e FROM x, y WHERE x.a = y.d)
SELECT t1.*, t2.* FROM t t1, t t2 WHERE t1.b = t2.e AND t1.a < 10 AND t2.a > 5
{code}
the plan may be as follow:
{code:java}
HashJoin(joinType=[InnerJoin], where=[=(b, e0)], select=[a, b, e, a0, b0, e0], build=[right])
:- Exchange(distribution=[hash[b]], shuffle_mode=[BATCH])
:  +- Calc(select=[a, b, e])
:     +- HashJoin(joinType=[InnerJoin], where=[=(a, d)], select=[a, b, d, e], build=[left])
:        :- Exchange(distribution=[hash[a]])
:        :  +- Calc(select=[a, b], where=[<(a, 10)])
:        :     +- TableSourceScan(table=[[default_catalog, default_database, x, source: [TestTableSource(a, b, c)]]], fields=[a, b, c])
:        +- Exchange(distribution=[hash[d]], reuse_id=[1])
:           +- Calc(select=[d, e])
:              +- TableSourceScan(table=[[default_catalog, default_database, y, source: [TestTableSource(d, e, f)]]], fields=[d, e, f])
+- Exchange(distribution=[hash[e]])
   +- Calc(select=[a, b, e])
      +- HashJoin(joinType=[InnerJoin], where=[=(a, d)], select=[a, b, d, e], build=[left])
         :- Exchange(distribution=[hash[a]])
         :  +- Calc(select=[a, b], where=[>(a, 5)])
         :     +- TableSourceScan(table=[[default_catalog, default_database, x, source: [TestTableSource(a, b, c)]]], fields=[a, b, c])
         +- Reused(reference_id=[1])
{code}
So I think it is useless to defined this option, only TABLE_OPTIMIZER_REUSE_SUB_PLAN_ENABLED will be ok.

 



--
This message was sent by Atlassian Jira
(v8.3.4#803005)