Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 25 Oct 2004 12:29:05 -0600
From:      Scott Long <scottl@freebsd.org>
To:        fandino@ng.fadesa.es
Cc:        freebsd-current@freebsd.org
Subject:   Re: FreeBSD 5.3b7and poor ata performance
Message-ID:  <417D45F1.9090504@freebsd.org>
In-Reply-To: <417D40A1.9030802@ng.fadesa.es>
References:  <14479.1098695558@critter.freebsd.dk> <417D25E8.6080804@ng.fadesa.es> <200410251928.01536.victor@alf.dyndns.ws> <200410251837.58257.Thomas.Sparrevohn@btinternet.com> <417D3F12.20302@DeepCore.dk> <417D40A1.9030802@ng.fadesa.es>

next in thread | previous in thread | raw e-mail | index | archive | help
fandino wrote:
> Søren Schmidt wrote:
> 
>>>>
>>>> please can you post how do you solved the problem?
>>>
>>>
>>>
>>> atacontrol mode <channel> DMA4 DMA4
>>> atacontrol mode <channel2> DMA4 DMA4
>>>
>>> The reason is that the motherboard only supports ATA100 on two out of 
>>> four IDE channels and hence you need to force the two other channels 
>>> to run at DMA66 
>>
>>
>>
>> How do you come to that conclusion ? There is no such limitation AFAIK.
>>
>> However you can only get a total sum of 133MB/s divided by number of 
>> disks (and minus some overhead on older system there is typically 
>> 110MB/s effective bandwidth)..
>>
>> The real explanation is much more likely that the timing specs are 
>> marginal (overclocked ?) for the disks, which causes problems..
> 
> 
> 
> no, it is my home PC in which I work so it's important stability
> (not overclocking) and disk redundancy (vinum, gmirror)

No, I think that he is saying that the ATA silicon is marginal and
probably overclocked by the vendor, not that you have overclocked your
CPU.

> 
> Also, there is an unresolvable question. Why two 52MB/s disks
> in raid0 has a throughput of 40MB/s and for raid1 18MB/s??

Would you _PLEASE_ stop trying to associate RAID with performance!
RAID is about reliability and reduncdancy, not about speed.  Some
cases can give you desirable performance increases as a side effect,
but that is not the primary goal.  Specifically in this case, the
GEOM raid classes are fairly new and have not had the benefit of
years of testing.  I'd much rather that the focus be on stability
and reliability for them, not speed.  Once the primary goals of
RAID are satisfied then we can start looking at performance.

Scott



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?417D45F1.9090504>