Re: postgresql random io test with 2 SSD Kingston V+100 500GB in (software) Raid1

Поиск
Список
Период
Сортировка
От Jeff
Тема Re: postgresql random io test with 2 SSD Kingston V+100 500GB in (software) Raid1
Дата
Msg-id E8B6180C-1CDA-49DA-BD89-DED82F49C25B@torgo.978.org
обсуждение исходный текст
Ответ на Re: postgresql random io test with 2 SSD Kingston V+100 500GB in (software) Raid1  (Laurent Laborde <kerdezixe@gmail.com>)
Ответы Re: postgresql random io test with 2 SSD Kingston V+100 500GB in (software) Raid1  (Laurent Laborde <kerdezixe@gmail.com>)
Список pgsql-performance
On Apr 19, 2011, at 8:49 AM, Laurent Laborde wrote:

> Write Mode: 10%
> Stallcheck at 1.000000
> Verbose
> Unknown units of blocks
> Arg: 1
> Read 1000000 blocks
> Added /home/ssd/big1
> 2.01%,  20122 read,   1978 written, 16097.57kB/sec 2012.20 iops
> 2.01%,      3 read,      0 written,    2.40kB/sec    0.30 iops
> 2.01%,      3 read,      0 written,    2.40kB/sec    0.30 iops
> 2.01%,      3 read,      0 written,    2.40kB/sec    0.30 iops
> 3.82%,  18036 read,   1779 written, 14428.73kB/sec 1803.59 iops
> 4.03%,   2175 read,    209 written, 1739.98kB/sec  217.50 iops
> 4.03%,      3 read,      0 written,    2.40kB/sec    0.30 iops
> 4.03%,      2 read,      0 written,    1.60kB/sec    0.20 iops
> 4.04%,      4 read,      0 written,    3.20kB/sec    0.40 iops
> 4.04%,      4 read,      0 written,    3.20kB/sec    0.40 iops

The performance here looks like the old jmicron based ssds that had
absolutely abysmal performance - the intel x25s do not suffer like
this. The x25's however suffer from the power durability Greg has
mentioned.  (And they will eventually need to be security erase'd to
restore performance - you'll start getting major write stalls). Looks
like you were on the cusp of stalling here.

btw, yay pgiosim! :)

--
Jeff Trout <jeff@jefftrout.com>
http://www.stuarthamm.net/
http://www.dellsmartexitin.com/




В списке pgsql-performance по дате отправления:

Предыдущее
От: Willy-Bas Loos
Дата:
Сообщение: not using partial index
Следующее
От: Tom Lane
Дата:
Сообщение: Re: not using partial index