Re: [gentoo-user] Intel(R) C600 SAS Controller

[thegeezer]

On 06/11/2014 11:49 AM, Stefan G. Weichinger wrote:
> Am 11.06.2014 12:41, schrieb thegeezer:
>
>>> everything around 380 MB/s ... only ~350 MB/s for
>>> /dev/vg01/winserver_disk0 (which still is nice)
>>
>> OK here is the clue.
>> if the LVs are also showing such fast speed, then please can you show
>> your command that you are trying to run that is so slow ?
> I originally noticed that virt-backup was slow so I looked into it and
> found some dd-command.
>
> My tests right now are like this:
>
>
>
> booze ~ # dd if=/dev/vg01/winserver_disk0 bs=1M   of=/dev/null
> ^C25+0 Datensätze ein
> 24+0 Datensätze aus
> 25165824 Bytes (25 MB) kopiert, 13,8039 s, 1,8 MB/s
>
> booze ~ # dd if=/dev/vg01/winserver_disk0 bs=4M   of=/dev/null
> ^C6+0 Datensätze ein
> 5+0 Datensätze aus
> 20971520 Bytes (21 MB) kopiert, 12,5837 s, 1,7 MB/s
>
> booze ~ # dd if=/dev/vg01/winserver_disk0    of=/dev/null
> ^C55009+0 Datensätze ein
> 55008+0 Datensätze aus
> 28164096 Bytes (28 MB) kopiert, 12,611 s, 2,2 MB/s
>
> So no "copy from-to same disk here" ... should be just plain reading, right?
>
> virt-backup does some ionice-stuff as well, but as you see, my
> test-commands don't.
>
> # cat /sys/block/sdc/queue/scheduler
> [noop] deadline cfq
>
> -> noop scheduler to let the controller do its own scheduling
>
>
>
> thanks, Stefan
>
>
>
yeah this is very very odd.
firstly there should not be such discrepancy between hdparm -t and dd if=
secondly you would imagine that the first dd would be cached and so
would be faster the second time round
please check for the turbo boost disable, i'll have a closer look at the
cgroups