Question

Postgresql 9.3

我有一张带有date_field的表格：

date_field timestamp without time zone

CREATE INDEX ix__table__date_field ON table USING btree (date_field)
    WHERE date_field IS NOT NULL;

然后我尝试使用我的部分索引：

EXPLAIN ANALYZE SELECT count(*) from table where date_field is not null;

Aggregate  (cost=29048.22..29048.23 rows=1 width=0) (actual time=41.714..41.714 rows=1 loops=1)
  ->  Seq Scan on table  (cost=0.00..28138.83 rows=363755 width=0) (actual time=41.711..41.711 rows=0 loops=1)
        Filter: (date_field IS NOT NULL)
        Rows Removed by Filter: 365583
Total runtime: 41.744 ms

但是它使用部分索引来比较日期：

EXPLAIN ANALYZE SELECT count(*) from table where date_field > '2015-1-1';

Aggregate  (cost=26345.51..26345.52 rows=1 width=0) (actual time=0.006..0.007 rows=1 loops=1)
  ->  Bitmap Heap Scan on table  (cost=34.60..26040.86 rows=121861 width=0) (actual time=0.005..0.005 rows=0 loops=1)
        Recheck Cond: (date_field > '2015-01-01 00:00:00'::timestamp without time zone)
        ->  Bitmap Index Scan on ix__table__date_field  (cost=0.00..4.13 rows=121861 width=0) (actual time=0.003..0.003 rows=0 loops=1)
              Index Cond: (date_field > '2015-01-01 00:00:00'::timestamp without time zone)
Total runtime: 0.037 ms

那么，为什么它不使用date_field is not null上的索引？

提前致谢！

Postgresql不使用部分索引

0 个答案: