sata to IDE - boot up issues

  • Thread starter Thread starter randalls
  • Start date Start date
R

randalls

Hi all, I originally ran my xp sp2 pc with a SATA main boot disk. Thi
failed, so I replaced it with a new IDE disk drive I had. this wa
some time ago, so I cant remember what I did at the time, but I hav
tried now to install a SATA data only drive. The problem is the PC i
trying to boot up from this sata drive & just goes dead as there is n
OS on this disk. I have checked the BIOS & it should be looking at th
IDE drive to boot up - but there is clearly a problem as soon as
enable the on board sata drive & install the disk, - an
suggestions/help ? thank
 
randalls said:
Hi all, I originally ran my xp sp2 pc with a SATA main boot disk. This
failed, so I replaced it with a new IDE disk drive I had. this was
some time ago, so I cant remember what I did at the time, but I have
tried now to install a SATA data only drive. The problem is the PC is
trying to boot up from this sata drive & just goes dead as there is no
OS on this disk. I have checked the BIOS & it should be looking at the
IDE drive to boot up - but there is clearly a problem as soon as I
enable the on board sata drive & install the disk, - any
suggestions/help ? thanks
Click to expand...

Some BIOS, have two levels to the boot order. At the top
level, you set up something like this.

1) Floppy
2) CDROM
3) Hard drive

That is a good general purpose boot order - as long as you don't have
a CD loaded, or a floppy loaded, then the system boots from a
hard drive.

The second part, is a menu of hard drives. Usually there are cursor
keys, or some other mechanism to move the disks around in the list.
So the second level, used to select hard drives, might look like this

Seagate 80GB 3800115A
Maxtor 40GB 7M3340
Samsung 20GB 1234321

The top hard drive, is the one it'll try to boot from.

Now, for such a BIOS, when you add a hard drive to the computer, it
usually screws up the second list. The drive you just added, may
have put itself at the top of the list. (This may have something
to do with the BIOS enumeration order, and which disk is discovered
first.)

Any time I add or remove a hard drive, I always enter the BIOS and
check that list.

Paul
 
If SATA was running before the IDE,then try rechecking the BIOS to allow
SATA & IDE to run together,by default its disabled.
 
This is somewhat similar to the problem I posted here on 20 May under "SATA
Western Digital......" except that my PC is trying to boot from a new SATA
drive instead of the existing 2 SATA drives in a RAID 0 array which have Win
XP and everything on them.

I've done everything Paul suggested without success.

Paul, you mention the Bios enumeration order and which disk is discovered
first. Can this be manipulated at all?

What would happen if I disconnect both SATA RAID drives and the new drive
and the SATA DVD-RAM which occupies the fourth SATA slot (and which worked
immediately on being installed on my MoBo) and rearrange them? Will any of
the 4 SATA slots do for anything or is there a correct order in which
particular peripherals (and particularly drives) should be connected, known
only to the MoBo (the manual doesn't mention this)? The company which
assembled my PC chose the slots to connect the 2 drives set up as a RAID
array. They happen to show up in BIOS as masters.The full details are in my
earlier post. I haven't touched those, although I've swapped the DVD-RAM
and new HD around, because I'm terrified I might mess up the RAID array
(which took me days to set up when I ftook delivery of the PC when new
and/or not be able to access the OS and/or lose data?

Is it risk freesafe or is better not to do this or would it make no
difference anyway?
John

P.S. Randall,
If like me, you are repeatedly connecting and dis-connecting the new drive
while trying to find the solution, make sure you recheck the HD priority in
the BIOS each time or the new drive will keep being listed as first!
 
John said:
This is somewhat similar to the problem I posted here on 20 May under
"SATA Western Digital......" except that my PC is trying to boot from a
new SATA drive instead of the existing 2 SATA drives in a RAID 0 array
which have Win XP and everything on them.

I've done everything Paul suggested without success.

Paul, you mention the Bios enumeration order and which disk is
discovered first. Can this be manipulated at all?
Click to expand...

Don't mind the following. Just making a few notes.

*******
GA-8i915P Duo Pro
http://www.gigabyte.com.tw/Products/Motherboard/Products_Overview.aspx?ProductID=1802

F6 BIOS contents -

8i9pdup.BIN
awardext.rom
ACPITBL.BIN
AwardBmp.bmp
awardeyt.rom
font1.awd
font2.awd
_EN_CODE.BIN
_B5_CODE.BIN
_GB_CODE.BIN
dbios.bmp
b169d.pxe Broadcom boot from net
SATARAID.BIN Intel(R) RAID for SATA - v4.0.0.6180
VIARAID.ROM VIA VT6410 RAID BIOS
PPMINIT.ROM
FINER.BIN XpressRecovery
dbf.bin Dual BIOS flash
SPECIAL.FNT

Southbridge -
1 IDE cable - managed by main BIOS
4 SATA ports - managed by SATARAID.BIN ? (no room in main BIOS display?)
VT6410
2 IDE cables - managed by VIARAID.ROM
*******

The enumeration order doesn't generally change.

Two things might control it. The first, is the order of execution
of BIOS modules. The main BIOS runs first. It definitely manages
the IDE cable on the Southbridge. Then, if the four SATA ports
are set to RAID mode, the SARARAID BIOS is loaded. It will do
detection on up to four disks. External chips run last. So the
VIARAID might run later than the Southbridge ports. It is really
hard to tell with the Gigabyte BIOS, because there should have
been room in the main disk display screen, for the max six disks
on the Southbridge.

On the PCI bus, you do have some control of enumeration order.
PCI slots are examined in address space order, and if the motherboard
designers did their job properly, PCI1 would be examined before
PCI5. People who have several SCSI cards, take advantage of
features like that, and juggle card order, in order to get
whatever property they're after. (Place card you expect to boot
from, in PCI1. In case there is insufficient room for option ROMs
to load.)

But, if you wanted to alter the relationship of PCI slots,
versus other hardware in the system, there is no way to
change that. At least on a desktop board - a server BIOS
might be different.

Things that are unclear in my mind, are the impact of disks
marked "active". I would have expected, that their status would
only matter at the instant you try to boot from them. But I don't
know if each BIOS module does some investigation during
disk detection or not. It is possible a USB mass storage
BIOS module, might do something like that. In the case
of the above BIOS, it is possible the USB mass storage code
is hiding in the main BIOS module. And it is unclear to me, when
the USB mass storage code would run - during main Southbridge
examination, or later.

What is somewhat puzzling, is how this could appear in your main
disk display screen in the BIOS. Quoted from your other thread:

"In Bios, the standard CMOS features show the following:

IDE Channel 0 Master HL-DTST DVD RAM GSH (The IDE DVD RW?)
IDE Channel 0 Slave None
IDE Channel 2 Master S_ATA1_ Raid drive
IDE Channel 2 Slave S_ATA2_ (the Western's id)
IDE Channel 3 Master S_ATA3_ Raid drive
IDE Channel 3 Slave S_ATA4_ (the Pioneer's id)"

I would not have expected that, based on the picture shown in the
Gigabyte manual. That looks more like other motherboard displays
and how they behave.

If this was my computer, I would

1) Boot into Linux (Knoppix or Ubuntu)
2) Use the "dd" (disk dump) command, to erase the first 50MB or so
of the new hard drive. This is so all partition information
that might have been placed on there at the factory, would be removed.

Alternatives, are erasing programs like DBAN (sourceforge.net), but
all other hard drives must be removed before running that one!
Very dangerous, as it erases all hard drives connected to the
computer, at the instant it is run. It can erase up to 99 disks
simultaneously.

You might also find, that if just the 400GB was connected,
you booted a Windows installer CD, that Windows might refuse
to install on the new disk. If that happened, I would again
suspect something at the very beginning of the disk, is
confusing the software. When that happened to me, that is
when I started reading about "dd". There is even a Windows
version of "dd", but I'm not comfortable using something
like that in Windows. You have to be absolutely sure about
the command syntax - unlike other Windows programs, that might
prevent you from damaging C:, the Windows version of "dd" would
have no trouble at all, erasing the very volume it is booted
from. Which is one of the reasons I've never used it (I've only
tried the --list option).

When booting from a Linux LiveCD, you cannot delete the CD, so
only the target disk can be accessed by "dd" in that case - as
long as only the disk to be erased is connected, it is
pretty hard to foul up the command syntax in a destructive
way. A command like this, from a Knoppix terminal window,
would erase the beginning of a hard drive at /dev/sda.
This should take less than 30 seconds.

sudo dd if=/dev/zero of=/dev/sda bs=512 count=10000

HTH,
Paul
 
That's interesting. I'm going to contact WD re what may be on the drive and
to see if they can throw any light on the problem before trying the Linux
route. I'm also waiting for a reply back from Gigabyte's technical
department but, from frustrated past experience, that might turn out to
darken the waters further.

I am right, I suppose in thinking that this is solely a SATA issue and that
no IDE related issues can be affecting the failure to get this WD drive
going? And the existing RAID 0 array set up on the first 2 drives installed
cannot have any effect on setting up a new drive, which although it is a
SATA drive, isn't part of the existing RAID array?

I'm also wondering if a reformat might provide the opportunity to deal with
all the drives? I take it that a reformat would mean setting up a RAID
array from the beginning on the 2 Hitachi drives and that there should be an
option to deal with the WD SATA drive at some stage?
Regards,

John
 
John said:
That's interesting. I'm going to contact WD re what may be on the drive and
to see if they can throw any light on the problem before trying the Linux
route. I'm also waiting for a reply back from Gigabyte's technical
department but, from frustrated past experience, that might turn out to
darken the waters further.

I am right, I suppose in thinking that this is solely a SATA issue and that
no IDE related issues can be affecting the failure to get this WD drive
going? And the existing RAID 0 array set up on the first 2 drives installed
cannot have any effect on setting up a new drive, which although it is a
SATA drive, isn't part of the existing RAID array?

I'm also wondering if a reformat might provide the opportunity to deal with
all the drives? I take it that a reformat would mean setting up a RAID
array from the beginning on the 2 Hitachi drives and that there should be an
option to deal with the WD SATA drive at some stage?
Regards,

John
Click to expand...

This is your setup, as I understand it.

*******
GA-8i915P Duo Pro
http://www.gigabyte.com.tw/Products/Motherboard/Products_Overview.aspx?ProductID=1802
http://america.giga-byte.com/FileList/Manual/motherboard_manual_8i915pduo(pro)_e.pdf

Southbridge ICH6R (max six drives, RAID mode used)

SATA - Hitachi 160GB SATA-150 (RAID0 array, F6 with RAID driver during install)
SATA - Hitachi 160GB SATA-150 (RAID0 array, F6 with RAID driver during install)
SATA - SATA Pioneer DVD RW
SATA - *New* Western Digital 400GB SATA SATA-300 jumpered to SATA-150
IDE1
Master - IDE DVD RW
Slave -

VIA VT6410 RAID IDE controller (a not very useful option, unless you have lots of IDE HDD)
IDE2
Master -
Slave -
IDE3
Master -
Slave -
*******

In the BIOS, your "SATA RAID/AHCI Mode" would be set to RAID. The "On-chip SATA Mode"
could be Auto or Enhanced - the fact that the above configuration performs correctly
without the WD 400 present, suggests whatever was used, is correct. Perhaps you
could check and record the settings, to see if there is anything "novel" about
how they were set up.

The ICH5 and ICH6 have some goofy terminology, used to handle their setup. Intel
wanted a way to support Win98, even if SATA drives were present. Their solution was
to allow four of six ports on the Southbridge, to be put in IDE emulation mode, so
that Win98 drivers could be fooled into thinking there were two ribbon cables for
the drives. The Gigabyte BIOS interface makes a convoluted way of selecting four
of six disks, and using legacy mode. (The Asus implementation for this, which I
can see on this computer, is no better, in that certain combinations of the
controls in the Asus BIOS, cause strange operating modes on the chipset for IDE
drives and slow operation.)

For WinXP, the Enhanced mode for the chipset, makes all six drives show up
as PCI space devices. WinXP SP1 has a native Microsoft driver which can handle
that, and that might be what is handling the SATA Pioneer drive right now.
Doing properties on the device, and checking the driver files, might tell you
something about that.

The new WD 400GB should be able to be handled by the Microsoft driver - I don't know
if the RAID driver has a "vanilla" mode, and is handling the WD 400, or it is
the Microsoft driver.

Your problem is at boot time, so something has convinced the BIOS to try to
boot from that new drive. The BIOS shouldn't really be looking at the drive as
a boot device, if the RAID array is first in the hard disk priority list.
But if it won't behave, then something needs to change on the WD 400GB contents.

If you don't like the Linux route, then disconnect the two Hitachi drives from
their SATA connectors, and use DBAN from Sourceforge. It will erase all drives
connected to the computer. You only want the WD 400GB connected, when booting
the DBAN floppy. (And yes, the DBAN forums have posts from people who ignored
that advice, and erased disks they didn't want erased!) There is also a
CD version available of some sort. (I've only used the floppy version myself.)

http://dban.sourceforge.net/

"Quick erase" might be the shortest option, but it will still take a while.
That is why I like the Linux "dd" option, as it is faster when wiping the
origin of the drive.

http://sourceforge.net/project/screenshots.php?group_id=61951&ssid=39777

Paul
 
Well, The Gods gave spoken. This arrived today from Gigabyte. I'll just
scratch around for a spare floopy then I'll give it a go. May end up doing
a complete reformat yet.

"Dear Sir,

Thank you for your kindly mail and supporting GIGABYTE.
About the issue you mentioned in your earlier mail, please try to get
preinstall driver for RAID controller from GIGABYTE website before OS
installation:
http://www.gigabyte.com.tw/Support/Motherboard/Driver_Model.aspx?ProductID=1802

You have to execute the exe file under Windows then save the files in a
blank floopy disk.

Then, follow OS installation procedure to install the driver properly.

Unfortunately, if the problem still remains, we suggest that you contact
your retailer/supplier, or, contact with local distributor for further help.
For their information, please kindly visit our website by following link:
http://www.gigabyte.com.tw/WhereToBuy/Default.aspx

At last, we apologize if our past service was not satisfactory. Due to the
vast amounts of questions we receive daily, we understand the inconvenience
of waiting for e-mail responses. If you still have any further question or
suggestion about our product, please do not hesitate to contact with us.
Thanks again for your understanding, and please receive our sincerely
apology.

******** Thank you for supporting GIGABYTE products ***********"
 
I've looked at the link Gigabyte sent and downloaded the file, opened it
onto a floppy and read the 'Read me'. It looks as if the OS will have to be
reinstalled, maybe even the existing RAID 0 configuration redone, just to
get this new SATA HD working. And I thought SATA was supposed to be simple
and easy?

John
 
Back
Top