Обсуждение: AW: [HACKERS] Some notes on optimizer cost estimates

Поиск
Список
Период
Сортировка

AW: [HACKERS] Some notes on optimizer cost estimates

От
Zeugswetter Andreas SB
Дата:
> As best I can measure on my hardware, the cost of a nonsequential
> disk read should be estimated at 4 to 5 times the cost of a sequential
> one --- 

Same experience here with RS6000 AIX hardware. 
450k/s random versus 4Mb/s sequential (16kb block readahead)
Compared with 256kb readahead the factor is even worse. 
Like 10Mb/s sequential.

Andreas