[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: BF24 and Promise Fasttrack problem



Hello,
Is there someone who had succes, using RAID1 with a Fasttrack controler.
OK, the installation succeeded (thanks to these messages ) and I'm only seeing 1 harddisk while using the server in a normal operation (/dev/ataraid/). When I unplug disk 2 (/dev/hdg) and reboot everything still works fine, but if disk 1(/dev/hde) is unplugged, and disk 2 is plugged (plugged in port 1 or 2 ), lilo is starting ,but soon giving up.

Messages like these apear : (by head, if you wanna need more, no probs)

printing eip : 00000000...
..
oops 0000
...
not tainted
..
bad eip value
kernel panic : Attempted to kill init


ps. If I mount the second disk (/dev/hdg), everything looks fine, the OS and data are nicely mirrored.

Someoneone has an idea or should I just give up on the fasttrack controler?

Hans



On Tuesday 07 March 2000 23:37, Yann Lollivier wrote:
Hi,

I tried it, it did  work...After a wee trick :

Actually I can see 10 different hex numbers, on five lines :

0xd000 [0xd007]
0xd400 [0xd403]
0xd800 [0xd807]
0xdc00 [0xdc03]
0xe000 [0xe03f]

I tried your magic formula with the first set of number (without brackets), and it just reduced my screen to a very narrow one...So I disabled 'set IRQ
for VGA'  in bios and it worked perfect !!!

Many thanks, I was going depressive about that...

By the way, what are those numbers exactly ? might be obvious for you but
I'm a kind of newbie...

Yann
IIRc they are IO adresses. I found it somewhere on the web (can't remember where, I remebered the solution) while having the same problem as you.
At 01:05 06/08/2002 +0000, Meir Kriheli wrote:
>On Thursday 01 August 2002 14:38, yann.lollivier@synteleya.com wrote:
> > Hi,
> >
> > I am trying without success to install the bf24 flavour on a server
> > containing a Promise Fasttrack 100 Raid Card. I've got 2 disk on linked
> > to the raid card, so I created a RAID1 array and enabled raid support
> > in BIOS. I cannot install debian as the installation process does not
> > detect any hard drive, however this card is supposed to be supported by
> > the 5.4 kernel...and is also apparently known to cause problems with
> > bf24. Anyone has a workaround on that ?
>
>Seems that bf24 cannot pick it autmatically (while Gentoo's kernel on the
>install cd picked it up with no problems). Here's what you should do:
>
>In a terminal window (alt+f2 or alt+f3 can't remember) run:
>
>cat /proc/pci
>
>and look for the info about the PDC controller. As part of the info you'll
>see
>5 hex numbers listed one after the other, let's name the (a)-(e)
>
>Write down the numbers, reboot the install cd again and at the boot prompt
>write:
>
>bf24 ide2=0x(a),0x(b+2) ide3=0x(c),0x(d+2)
>
>b+2 means the value of b incremented by 2, and same applies to d+2.
>
>Now the controller and raid should be detected.
>
>This is from memory, so tell me how it goes
>

To be complete, I had to add the string, append="bf24 ide2=0x(a),0x(b+2) ide3=0x(c),0x(d+2)" in my lilo.conf as well (/target/etc/lilo.conf) to be able to boot from harddisk.


> > More, Promise FastTack is not a real HW Raid, does someone knows how it
> > compares to software Raid (performance, fault tolerance...) ? If it's
> > as good I might go for it..
> >
> > Regards,
> > Yann
>
>I was very disappointed by it. While working I pulled the plug on one of
> the hd's (RAID1 config) to simulate a dying HD, lot of timeout message
> started appearing and the systems froze. Once restarted the system didn't
> boot.
>
>I bought a real IDE Raid soultion (LSI Logic), and it is awesome (although > a bit expansive). It handled to dying hd trick with no problems, after a > short pause the system continued working normally. When re-connected the
> power it started rebuilding the array automatically, I was very pleased
> by it.
>
>HTH
>--
>Meir Kriheli
>
>
>--
>To UNSUBSCRIBE, email to debian-user-request@lists.debian.org
>with a subject of "unsubscribe". Trouble? Contact
> listmaster@lists.debian.org
--
Meir Kriheli



Reply to: