Partition Iterator算子消除

场景描述

在当前分区表架构中，执行器通过Partition Iterator算子去迭代访问每一个分区。当分区剪枝结果只有一个分区时，Partition Iterator算子已经失去了迭代器的作用，在此情况下消除Partition Iterator算子，可以避免执行时一些不必要的开销。由于执行器的PIPELINE架构，Partition Iterator算子会重复执行，在数据量较大的场景下消除Partition Iterator算子的收益十分可观。

示例

消除Partition Iterator算子在GUC参数partition_iterator_elimination开启后才能生效，示例如下：

gaussdb=# 
CREATE TABLE test_range_pt (a INT, b INT, c INT)
PARTITION BY RANGE (a)
(
    PARTITION p1 VALUES LESS THAN (2000),
    PARTITION p2 VALUES LESS THAN (3000),
    PARTITION p3 VALUES LESS THAN (4000),
    PARTITION p4 VALUES LESS THAN (5000),
    PARTITION p5 VALUES LESS THAN (MAXVALUE)
)ENABLE ROW MOVEMENT;

gaussdb=# EXPLAIN SELECT * FROM test_range_pt WHERE a = 3000;
                                    QUERY PLAN                                    
----------------------------------------------------------------------------------
 Partition Iterator  (cost=0.00..25.31 rows=10 width=12)
   Iterations: 1
   ->  Partitioned Seq Scan on test_range_pt  (cost=0.00..25.31 rows=10 width=12)
         Filter: (a = 3000)
         Selected Partitions:  3
(5 rows)

gaussdb=# SET partition_iterator_elimination = on;
SET
gaussdb=# EXPLAIN SELECT * FROM test_range_pt WHERE a = 3000;
                                 QUERY PLAN                                 
----------------------------------------------------------------------------
 Partitioned Seq Scan on test_range_pt  (cost=0.00..25.31 rows=10 width=12)
   Filter: (a = 3000)
   Selected Partitions:  3
(3 rows)

gaussdb=# DROP TABLE test_range_pt;