案例：增加JOIN列非空条件

  1. SELECT * FROM join_a a JOIN join_b b ON a.b = b.b;

执行计划下：

  1.                                                       QUERY PLAN
  2. ----------------------------------------------------------------------------------------------------------------------
  3.  Hash Join  (cost=58.35..14677.69 rows=1074607 width=16) (actual time=23.374..23.384 rows=10 loops=1)
  4.    Hash Cond: (a.b = b.b)
  5.    ->  Seq Scan on join_a a  (cost=0.00..2248.10 rows=100010 width=8) (actual time=0.495..12.551 rows=100010 loops=1)
  6.    ->  Hash  (cost=31.49..31.49 rows=2149 width=8) (actual time=0.614..0.614 rows=1000 loops=1)
  7.           Buckets: 32768  Batches: 1  Memory Usage: 40kB
  8.          ->  Seq Scan on join_b b  (cost=0.00..31.49 rows=2149 width=8) (actual time=0.009..0.183 rows=1000 loops=1)
  9.  Total runtime: 23.716 ms
  10. (7 rows)

优化分析

  1. 分析执行计划可知，在顺序扫描阶段耗时较多。

  2. 建议在语句中手动添加JOIN列的非空判断，修改后的语句如下所示。

    1. SELECT
    2.  * 
    3. SELECT * FROM join_a a JOIN join_b b ON a.b = b.b where a.b IS NOT NULL;

    执行计划如下：

    1.                                                      QUERY PLAN
    2. ---------------------------------------------------------------------------------------------------------------------
    3.  Hash Join  (cost=58.22..14560.97 rows=1063762 width=16) (actual time=13.237..13.247 rows=10 loops=1)
    4.    Hash Cond: (a.b = b.b)
    5.    ->  Seq Scan on join_a a  (cost=0.00..2248.10 rows=99510 width=8) (actual time=12.417..12.422 rows=10 loops=1)
    6.          Filter: (b IS NOT NULL)
    7.          Rows Removed by Filter: 100000
    8.    ->  Hash  (cost=31.49..31.49 rows=2138 width=8) (actual time=0.566..0.566 rows=1000 loops=1)
    9.           Buckets: 32768  Batches: 1  Memory Usage: 40kB
    10.          ->  Seq Scan on join_b b  (cost=0.00..31.49 rows=2138 width=8) (actual time=0.011..0.229 rows=1000 loops=1)
    11.                Filter: (b IS NOT NULL)
    12.  Total runtime: 13.556 ms
    13. (10 rows)