Tom said:
When performing an RIS install of WinXP Pro from a Win 2k3 RIS server the
boot.ini file gets created with the signature(xxx) syntax rather than the
multi(0) syntax. Setup from CD works without any problems. Anyone else
experienced this problem before?
I'm having the same problem right now only with a win2k3 install not
XP. I've posted this on a forum:
I'm trying to install W2k3 via RIS on some Sun X4100 rackmounts. These
have got LSI Logic 1064 serial SCSI controllers in them and are running
AMD64 Opterons. After the RIS remotely installs the OS, upon reboot I
get this message:
"Windows could not start because of a computer disk hardware
configuration problem. Could not read from selected boot disk. Check
boot path and disk hardware. Please check the Windows documentation
about hardware disk configuration and your hardware disk configuration
and your hardware reference manuals for additional information. Boot
Failed."
Basically, the only things added to the SIF file are the drivers for
the SAS disks. I have 10 machines to install this on and for 2 of them
there were no problems (identical hardware) which is why its also
confusing that it doesnt work on the others.
This happened each time I performed a RIS installation. I assumed that
it
was a hardware problem so I ran it on two other servers with the same
results.
After some internet searching I found hints that it might be something
to do with the boot.ini and booting up with a Linux bootcd I found that
the boot.ini looked like this:
[boot loader]
timeout=30
Default= signature(c400c400)disk(2)rdisk(0)partition(1)\Windows
[operating systems]
signature(3892r94)disk(0)rdisk(0)partition(1)\WINDOWS="Windows Server
2003 Enterprise Edition"
Rather than showing "multi(0)...\Windows", its using
signature(c400c400) and after a further search I found the following MS
kb article:
http://support.microsoft.com/?kbid=227704. This indicated a
problem with INT13 which just isnt the case and the SCSI bios is not
disabled. There is only one drive in there so I've no idea where the
disk(2) came from. The solutions I can think of to temporarily get
around this is to use a bootdisk and manually edit the boot.ini or use
RiPrep from one of the successful installs to use as an image for the
others. Either way is certainly not a good solution.
Can anyone help with this and force the RIS to make a boot.ini with
multi(0)disk(0) instead of signature(c400c400)disk(2)?
