[nas-2000] Re: Performance SMB vs FTP
- From: philipp Wehrheim <flipstar@xxxxxxx>
- To: nas-2000@xxxxxxxxxxxxx
- Date: Thu, 17 May 2007 19:05:42 +0200
one thing that is different is that i don't
use dropbear to connect to the nas but the serial cable
guess dropbear takes quite some power.
do you get better results without being
connected via ssh?
Aurelien schrieb:
> Excatly the same....
> I don't understand why we can get so much difference... ( :'( :'( :'( :'( )
>
> root@NASDRIVE:~ # cat /proc/cpuinfo
> Processor : FA52Xid(wb) rev 1 (v4l)
> BogoMIPS : 69.83
> Features : swp half thumb
> CPU implementer : 0x66
> CPU architecture: 4
> CPU variant : 0x0
> CPU part : 0x526
> CPU revision : 1
> Cache type : write-back
> Cache clean : cp15 c7 ops
> Cache lockdown : format B
> Cache format : Harvard
> I size : 8192
> I assoc : 2
> I line length : 16
> I sets : 256
> D size : 4096
> D assoc : 2
> D line length : 16
> D sets : 128
>
> Hardware : Sword sl2312
> Revision : 0000
> Serial : 0000000000000000
>
>> This is what my cpuinfo says ...
>>
>> What do you get?
>>
>> root@icybox:~ # cat /proc/cpuinfo
>> Processor : FA52Xid(wb) rev 1 (v4l)
>> BogoMIPS : 69.83
>> Features : swp half thumb
>> CPU implementer : 0x66
>> CPU architecture: 4
>> CPU variant : 0x0
>> CPU part : 0x526
>> CPU revision : 1
>> Cache type : write-back
>> Cache clean : cp15 c7 ops
>> Cache lockdown : format B
>> Cache format : Harvard
>> I size : 8192
>> I assoc : 2
>> I line length : 16
>> I sets : 256
>> D size : 4096
>> D assoc : 2
>> D line length : 16
>> D sets : 128
>>
>> Hardware : Sword sl2312
>> Revision : 0000
>> Serial : 0000000000000000
>>
>>
>> Aurelien schrieb:
>>
>>> Maybe that NAS-1000 and NAS-2000 do not get the same CPU???
>>> That could also explain the cooler and performances....
>>>
>>> philipp Wehrheim a écrit :
>>>
>>>> Thats strange specialy because you have the NAS1000 right?
>>>>
>>>> and so you don't a fan on the NAS ...
>>>>
>>>> Aurelien schrieb:
>>>>
>>>>
>>>>> I've tried to see if there were any difference between hardware of my
>>>>> NAS and the Flip's one.
>>>>> I've seen that Flip got a thermal cooler on his NAS CPU. There is
>>>>> nothing on mine (see attached pictures).
>>>>> I will try to see if the unit CPU is really hot during transfer. That
>>>>> could explain, for a same firmware, the hugh performance
>>>>> difference that we've got.
>>>>>
>>>>> philipp Wehrheim a écrit :
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> not really guess it would be really usefull
>>>>>> to start all daemons that support it by the inetd
>>>>>>
>>>>>> Julius Loman schrieb:
>>>>>>
>>>>>>
>>>>>>> On Thursday 17 May 2007 16:28, Tom Haukap wrote:
>>>>>>>
>>>>>>>> Triggerd by the recent discussion about ftp performance I check by
>>>>>>>> box and
>>>>>>>> found out that the SMB transfer rate are only half of the once from
>>>>>>>> ftp.
>>>>>>>>
>>>>>>>> I get nearly 7 MB/s from FTP but only around 3.5 MB/s using SMB.
>>>>>>>>
>>>>>>> you are lucky having such values!!
>>>>>>> i'm getting 1.4MB/s with SMB and 2.8 with CIFS
>>>>>>>
>>>>>>> i've also done some testing and compiled another smbd in gentoo
>>>>>>> chroot on NAS with same performance results! so now I think the
>>>>>>> performance will not be much better via smb/cifs protocol on this
>>>>>>> NAS
>>>>>>> hardware.
>>>>>>>
>>>>>>> at least this cpu has ~70 bogomips compared to ~4000 with my 2.0GHz
>>>>>>> Pentium-M
>>>>>>>
>>>>>>> when downloading from NAS, arm cpu in NAS is very busy for smbd (95%
>>>>>>> or more) and when downloading from my laptop to another computer it
>>>>>>> is around 2% (at laptop).
>>>>>>> so now i guess the performance of SMB is limited with NAS hardware.
>>>>>>> has anyone got better results than yours 3.5MB/s ?
>>>>>>>
>>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------
>>>>>
>>>>>
>>>>>
>>>>
>>>
>>
>>
>
>
Other related posts:
