HPT372 on KR7A-133R (ATA133) on production server

HPT372 on KR7A-133R (ATA133) on production server

[Vasja J Zupan]

Hi,
I'm buying a production server for webservices for mobile operators.

My HW provider suggested Abit's KR7A-133R (ATA133)  with HPT372.

Is this motherboard already fully supported and when will stable kernel be
ready for production use on these motherboards? (btw - any advice on good
amd combo is appreciated)

Thank you for your help and I'd appreciate a personal cc: cos I'm not a
member of this list.

Thanx!
Vasja

Re: HPT372 on KR7A-133R (ATA133) on production server

[Andy Jeffries]

On Thu, 18 Apr 2002 16:16:47 +0200, "Vasja J Zupan" <vasja@nuedi.com>
wrote:
> Hi,
> I'm buying a production server for webservices for mobile operators.
> 
> My HW provider suggested Abit's KR7A-133R (ATA133)  with HPT372.
> 
> Is this motherboard already fully supported and when will stable kernel
> be ready for production use on these motherboards? (btw - any advice on
> good amd combo is appreciated)
> 
> Thank you for your help and I'd appreciate a personal cc: cos I'm not a
> member of this list.

It was crashing the Kernel, I helped a friend get his working with a patch
which he has uploaded on his website.  To date, I haven't had a reply when
I tried to ask who to submit it to on here (2.4.18 isn't fixed AFAIK).

http://www.timj.co.uk/linux/

Cheers,




-- 
Andy Jeffries
Linux/PHP Programmer
http://www.andyjeffries.co.uk/

- Windows Crash HOWTO: compile the code below in VC++ and run it!
main (){for(;;){printf("Hung up\t\b\b\b\b\b\b");}}

Re: HPT372 on KR7A-133R (ATA133) on production server

[Vasja J Zupan]

Thanks!!!!
Can you tell me if it was crashing only if raid controller was used, or did
it crash on normal ide too?
I could wait until stable kernel is out, living only with software raid and
use HW afterwards. or use your patch of course!

Does this motherboard have any other (Linux or any other) performance and
stability issues?

Rgds,
Vasja

----- Original Message -----
From: "Andy Jeffries" <lkml@andyjeffries.co.uk>
To: "Vasja J Zupan" <vasja@nuedi.com>
Cc: <linux-kernel@vger.kernel.org>
Sent: Thursday, April 18, 2002 5:12 PM
Subject: Re: HPT372 on KR7A-133R (ATA133) on production server


> On Thu, 18 Apr 2002 16:16:47 +0200, "Vasja J Zupan" <vasja@nuedi.com>
> wrote:
> > Hi,
> > I'm buying a production server for webservices for mobile operators.
> >
> > My HW provider suggested Abit's KR7A-133R (ATA133)  with HPT372.
> >
> > Is this motherboard already fully supported and when will stable kernel
> > be ready for production use on these motherboards? (btw - any advice on
> > good amd combo is appreciated)
> >
> > Thank you for your help and I'd appreciate a personal cc: cos I'm not a
> > member of this list.
>
> It was crashing the Kernel, I helped a friend get his working with a patch
> which he has uploaded on his website.  To date, I haven't had a reply when
> I tried to ask who to submit it to on here (2.4.18 isn't fixed AFAIK).
>
> http://www.timj.co.uk/linux/
>
> Cheers,
>
>
>
>
> --
> Andy Jeffries
> Linux/PHP Programmer
> http://www.andyjeffries.co.uk/
>
> - Windows Crash HOWTO: compile the code below in VC++ and run it!
> main (){for(;;){printf("Hung up\t\b\b\b\b\b\b");}}
>

Re: HPT372 on KR7A-133R (ATA133) on production server

[Vincent Bernat]

OoO Lors de la soirée naissante du jeudi 18 avril 2002, vers 17:12,
Andy Jeffries <lkml@andyjeffries.co.uk> disait:

> It was crashing the Kernel, I helped a friend get his working with a patch
> which he has uploaded on his website.  To date, I haven't had a reply when
> I tried to ask who to submit it to on here (2.4.18 isn't fixed
> AFAIK).

I was experiencing DMA trouble with 2.4.18 when using software RAID 0
on this motherboard. There was no problem when not using RAID
software, even when doing parallel transfer between two disks. The
problem disappeared with Alan Cox' 2.4.19-pre4-ac6 (not sure, will
check the version, but the problem still was here with
2.4.19-pre2-ac4).

There was no problem when using software RAID for the card (but
performances were poor). But definitly, it is still under heavy
development.
-- 
 /*
  * Hash table gook..
  */
	2.4.0-test2 /usr/src/linux/fs/buffer.c

Re: HPT372 on KR7A-133R (ATA133) on production server

[Nicolae P. Costescu]

Vasja
I evaluated this board for linux server use. Stay away from it.
When you hook up and install the hpt raid, it corrupts disk files. This is 
100% reproducible.
The non hpt controller on board works ok.
I believe this is related to the via chipset problem with the PCI bus 
(doesn't allow bursts that are long enough for high bandwidth pci devices).

We went with i845 based Intel 845BG boards and p4/1.8GHz cpu. Cost 
difference is not much and performance was much better
If you like an athlon solution, Tyan's 2466N-4M AMD MPX based board works 
well too.

You can read about the via bug here...

http://www.tecchannel.de/hardware/817/index.html

I had this problem with KT133A and KT266A motherboards.

Email me if you want more info.
Thanks
Nick


At 04:16 PM 4/18/2002 +0200, Vasja J Zupan wrote:
>Hi,
>I'm buying a production server for webservices for mobile operators.
>
>My HW provider suggested Abit's KR7A-133R (ATA133)  with HPT372.
>
>Is this motherboard already fully supported and when will stable kernel be
>ready for production use on these motherboards? (btw - any advice on good
>amd combo is appreciated)
>
>Thank you for your help and I'd appreciate a personal cc: cos I'm not a
>member of this list.
>
>Thanx!
>Vasja
>
>-
>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at  http://www.tux.org/lkml/

****************************************************
Nicolae P. Costescu, Ph.D.  / Senior Developer
Stronghold Technologies
46040 Center Oak Plaza, Suite 160 / Sterling, Va 20166
Tel: 571-434-1472 / Fax: 571-434-1478

Re: HPT372 on KR7A-133R (ATA133) on production server

[Etienne Lorrain]

  I have this exact motherboard, with CRUCIAL 256MB 184DIMM PC2100 NP CL2.5
 memory.
 You first have to patch your kernel like Andy Jeffries said to activate
 the RAID controller. Most of the distribution still did not do it,
 you probably have to compile your own kernel.

 For a long time, I have only been able to run it 100% CPU for one hour,
 no disk access, if I set the memory clock to 100/200 MHZ.

 I am currently experimenting running at 133/266 MHz with the two
 parameters "DDR DQS Input Delay" and "DDR DQS Output Delay" changed
 to manual and increased by 2.
 Their autodetection does not seem to work, I am not sure.
 It seems stable - no error noticed for the last few weeks.
 Note that 100/200 to 133/266 did not seem to increase the CPU speed
 a lot, my usual task takes 39 minutes 100% CPU in 133/266 and
 40 min in 100/200 mode.
 I am not on a production server, so your millage may vary.

  Just my €0.02,
  Etienne.

___________________________________________________________
Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français !
Yahoo! Mail : http://fr.mail.yahoo.com

Re: HPT372 on KR7A-133R (ATA133) on production server

[Vasja J Zupan]

Ok,
I see, what CPU do you use? I'm planning to go for Athlon XP 1800+.
I do not see my services would use 100% CPU at any time but I cannot be 100%
certain!
I'll add 512MB DDR PC266 RAM (vendor unknown at this time).

I also plan to add raiserfs for two mirror 40GB WD ata100 7200rpm.

Any suggestions?

rgds,
Vasja

----- Original Message -----
From: "Etienne Lorrain" <etienne_lorrain@yahoo.fr>
To: <linux-kernel@vger.kernel.org>
Cc: <vasja@nuedi.com>
Sent: Thursday, April 18, 2002 5:59 PM
Subject: Re: HPT372 on KR7A-133R (ATA133) on production server


>   I have this exact motherboard, with CRUCIAL 256MB 184DIMM PC2100 NP
CL2.5
>  memory.
>  You first have to patch your kernel like Andy Jeffries said to activate
>  the RAID controller. Most of the distribution still did not do it,
>  you probably have to compile your own kernel.
>
>  For a long time, I have only been able to run it 100% CPU for one hour,
>  no disk access, if I set the memory clock to 100/200 MHZ.
>
>  I am currently experimenting running at 133/266 MHz with the two
>  parameters "DDR DQS Input Delay" and "DDR DQS Output Delay" changed
>  to manual and increased by 2.
>  Their autodetection does not seem to work, I am not sure.
>  It seems stable - no error noticed for the last few weeks.
>  Note that 100/200 to 133/266 did not seem to increase the CPU speed
>  a lot, my usual task takes 39 minutes 100% CPU in 133/266 and
>  40 min in 100/200 mode.
>  I am not on a production server, so your millage may vary.
>
>   Just my ?0.02,
>   Etienne.
>
> ___________________________________________________________
> Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français !
> Yahoo! Mail : http://fr.mail.yahoo.com
>