Postgres 查询计划为什么行估计如此错误

如何解决Postgres 查询计划为什么行估计如此错误

Postgres 版本：12

查询： 解释（分析为真、详细为真、成本为真、缓冲区为真、时间为真） SELECT MIN("id"),MAX("id") FROM "public"."hotel_slot_inventory" WHERE ( "updated_at" >= '2021-03-02 13:30:03' AND "updated_at"

查询计划：

 Result  (cost=512.44..512.45 rows=1 width=8) (actual time=57839.244..57839.250 rows=1 loops=1)
   Output: $0,$1
   Buffers: shared hit=1 read=454374 written=185
   I/O Timings: read=54564.571 write=2.686
   InitPlan 1 (returns $0)
     ->  Limit  (cost=0.57..256.22 rows=1 width=4) (actual time=57599.761..57599.764 rows=1 loops=1)
           Output: hotel_slot_inventory.id
           Buffers: shared read=453546 written=185
           I/O Timings: read=54330.640 write=2.686
           ->  Index Only Scan using hotel_slot_inventory_id_updated_at_idx on public.hotel_slot_inventory  (cost=0.57..3285663.29 rows=12852 width=4) (actual time=57599.758..57599.759 rows=1 loops=1)
                 Output: hotel_slot_inventory.id
                 Index Cond: ((hotel_slot_inventory.id IS NOT NULL) AND (hotel_slot_inventory.updated_at >= '2021-03-02 13:30:03'::timestamp without time zone) AND (hotel_slot_inventory.updated_at < '2021-03-03 06:15:19.127884'::timestamp without time zone))
                 Heap Fetches: 0
                 Buffers: shared read=453546 written=185
                 I/O Timings: read=54330.640 write=2.686
   InitPlan 2 (returns $1)
     ->  Limit  (cost=0.57..256.22 rows=1 width=4) (actual time=239.468..239.470 rows=1 loops=1)
           Output: hotel_slot_inventory_1.id
           Buffers: shared hit=1 read=828
           I/O Timings: read=233.931
           ->  Index Only Scan Backward using hotel_slot_inventory_id_updated_at_idx on public.hotel_slot_inventory hotel_slot_inventory_1  (cost=0.57..3285663.29 rows=12852 width=4) (actual time=239.465..239.465 rows=1 loops=1)
                 Output: hotel_slot_inventory_1.id
                 Index Cond: ((hotel_slot_inventory_1.id IS NOT NULL) AND (hotel_slot_inventory_1.updated_at >= '2021-03-02 13:30:03'::timestamp without time zone) AND (hotel_slot_inventory_1.updated_at < '2021-03-03 06:15:19.127884'::timestamp without time zone))
                 Heap Fetches: 0
                 Buffers: shared hit=1 read=828
                 I/O Timings: read=233.931
 Planning Time: 10.577 ms
 Execution Time: 57839.332 ms
(28 rows)

在两个 InitPlan 中，rows=12852 而实际 rows=1。这是为什么？ Index only scan后单独添加了limit子句。

编辑索引膨胀详细信息：

真实大小：3751411712 = 3.49 GB

额外大小：470237184 = 448 MB

额外比率：12.53

填充因子：90

膨胀大小：107053056 = 102 MB

膨胀比率：2.85

表格膨胀大小： 膨胀大小：475283456 = 453 MB

膨胀比率：5.088

解决方法

看起来索引膨胀变形了。试试

REINDEX INDEX hotel_slot_inventory_id_updated_at_idx;

或

VACUUM (FULL) hotel_slot_inventory;

,

这是正常的。 “仅索引扫描”行的预期行数给出了预期满足索引条件的行数，如果扫描完成。实际行反映了 LIMIT 使其提前停止的事实。请注意，LIMIT 的成本估算低于索引扫描的成本估算，因为该估算也是针对完整扫描给出的，然后由 LIMIT “回退”。