Re: [HACKERS] Sequential scan speed, mmap, disk i/o
| От | Thomas G. Lockhart |
|---|---|
| Тема | Re: [HACKERS] Sequential scan speed, mmap, disk i/o |
| Дата | |
| Msg-id | 355CFC8E.91BF31DF@alumni.caltech.edu обсуждение исходный текст |
| Ответ на | Re: [HACKERS] Sequential scan speed, mmap, disk i/o (Bruce Momjian <maillist@candle.pha.pa.us>) |
| Ответы |
Re: [HACKERS] Sequential scan speed, mmap, disk i/o
Re: [HACKERS] Sequential scan speed, mmap, disk i/o |
| Список | pgsql-hackers |
> Basically, Linux is double my speed for 8k mmap'ed chunks. Around 32k
> chunks, I get closer, and 8mb chunks are the same. Glad to hear Linux
> has optimized mmap() recently, because BSD/OS looks much slower than
> Linux on this.
Well Bruce, don't be too happy. Most people aren't yet running the
optimized kernel; don't know if any of the benchmarks came from someone
running a bleeding-edge development version, which is what 2.1.99 would
be; first feature-freeze release in preparation for v2.2 afaik :)
And scrappy, no need to note that _all_ Linux kernels are bleeding edge
releases :)
- Tom (from his Linux box...)
В списке pgsql-hackers по дате отправления: