[GENERAL] Why so long?

Hello,

I am confused. I have a table that has an incrementing primary key id.

When I select max(id) from table is returns almost instantly but
when I select min(id) from table it takes longer than I want to wait.

Shouldn't postgresql be able to quickly find the minimum id value in the index?


pmacct=# explain select max(id) from netflow;
                                                        QUERY PLAN                                                       
--------------------------------------------------------------------------------------------------------------------------
 Result  (cost=1.13..1.14 rows=1 width=0)
   InitPlan 1 (returns $0)
     ->  Limit  (cost=0.71..1.13 rows=1 width=8)
           ->  Index Only Scan Backward using netflow_pkey on netflow  (cost=0.71..3799108784.10 rows=9123246080 width=8)
                 Index Cond: (id IS NOT NULL)
(5 rows)

pmacct=# explain select min(id) from netflow;
                                                   QUERY PLAN                                                   
-----------------------------------------------------------------------------------------------------------------
 Result  (cost=1.13..1.14 rows=1 width=0)
   InitPlan 1 (returns $0)
     ->  Limit  (cost=0.71..1.13 rows=1 width=8)
           ->  Index Only Scan using netflow_pkey on netflow  (cost=0.71..3799108784.10 rows=9123246080 width=8)
                 Index Cond: (id IS NOT NULL)
(5 rows)

pmacct=# \timing
Timing is on.
pmacct=# select max(id) from netflow;
     max    
-------------
 17547256873
(1 row)

Time: 0.626 ms