I hope people can follow this description...

I just took a sustem that was running rhel4 and put down an image from 
rhel3.  The installation completed and the system rebooted.  For grins, 
I ran parted to look at the partition table for /dev/cciss/c0d0 and saw 
this:

[EMAIL PROTECTED] root]# parted /dev/cciss/c0d0
GNU Parted 1.6.3
Copyright (C) 1998, 1999, 2000, 2001, 2002 Free Software Foundation, Inc.
This program is free software, covered by the GNU General Public License.

This program is distributed in the hope that it will be useful, but 
WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or 
FITNESS FOR A
PARTICULAR PURPOSE.  See the GNU General Public License for more details.

Using /dev/cciss/c0d0
Error: The partition table on /dev/cciss/c0d0 is inconsistent.  There 
are many
reasons why this might be the case.  However, the most likely reason is that
Linux detected the BIOS geometry for /dev/cciss/c0d0 incorrectly.  GNU 
Parted
suspects the real geometry should be 8854/255/63 (not 17433/255/32).  
You should
check with your BIOS first, as this may not be correct.  You can inform 
Linux by
adding the parameter cciss/c0d0=8854,255,63 to the command line.  See 
the LILO
or GRUB documentation for more information.  If you think Parted's suggested
geometry is correct, you may select Ignore to continue (and fix Linux 
later).
Otherwise, select Cancel (and fix Linux and/or the BIOS now).
Ignore/Cancel? c
Information: The operating system thinks the geometry on /dev/cciss/c0d0 is
17433/255/32.  Therefore, cylinder 1024 ends at 4079.999M.

I then decided to try putting back rhel4 on it and got the same errors 
as before when I tried to install rhel4 on top of a system that used to 
have rhel3.  Therefore, the rhel3 installation is obviously configuring 
the disk in a way that not only causes parted to print the above 
warning, it also prevents one from installing a valid golden image that 
was taken of a rhel4 system.  I'm not convinced so much that it's an 
rhel3 vs rhel4 thing but there is clearly something different between 
the two...

clearly a question for a partitioning heavy, but does SI completely 
remove all partitions before it lays down the new one?  Is there an 
additional level of removal that it might be able to do to assure that 
it will be successful in recreating the partitions?

In any event I now have 2 system, both which used to run rhel4 and not 
neither can any more and can only run rhel3.  I'm hoping someone will 
come up with a solution that will allow me to put rhel4 back on them as 
opposed to having to manually install from disk (that's so 1990s)...

-mark

Mark Seger wrote:

> Ufortunately this is going to be complicated to explain, but I'll try 
> to be coherent...
>
> I installed rhel4 onto a single cciss disk (on a machine that has 2 of 
> them) and took an image.  I then successfully installed ithat image on 
> 6 other machines multiple times without error.
>
> I then wanted to see if SI could deal with rhel3 which has an older 
> version of grub that I know to have problems.  The first odd thing 
> about the installation was it complained about the disk geometery, 
> claming something was inconsistent and did I want to ignore it (feels 
> like this was caused by SI).  I said yes, assuming the installation 
> process could deal with it but I also manually deleted/recreated all 
> the partitions just to have a known quantity.  I gave the first one 
> the label /boot and made it 100MB.  The second I defined as swap and 
> made it 2048MB and the third as / telling it to fill the available space.
>
> The one curious thing was that when I looked at the disk layout a 
> small chunk of free space was inserted in front of the boot partitions 
> and so /boot didn't start at the beginning.  In any event, the system 
> built/booted correctly.  I took an image with system imager and was 
> able to reimage the target system and boot it.
>
> Next, I tried to put the rhel4 system back on top of the system than 
> had the rhel3 image on it.  The resultant system wouldn't boot and 
> actually hung in the middle of the process.  I took a screen shot and 
> can forward it if anyone cares.  But perhaps more important, I tried 
> again to put down a rhel3 image and it succeeded and booted.  I went 
> back and tried to reload rhel4 but this time with an exit statement in 
> the autoinstall script so I could see any errors that were generated.  
> Since I didn't know how to capture them in a file, here they are as 
> typed in by me:
>
> Probing devices to guess BIOS drives.  This may take a long time.
> end_requesst: I/O error, dev fd0, sector 0
> [repeated 14 times]
> Installation finished.  No errors reported.
> This is the contents of the device map /boot/grub/device.map.
> Check if this is correct or not.  If any of teh lines is incorrect,
> fix it and re-run the script 'grub-install'.
>
> (fd0)    /dev/fd0
> (hd1)    /dev/cciss/c0d0
> (hd2)    /dev/cciss/c0d1
> WARNING: Label SW-cciss/c0d0p2 not found anywhere on the system!  at 
> /usr/lib/sysconfig/Boot/Grub.pm line 207
> Probing devices to guess BIOS drives.  This may take a long time.
> end_requesst: I/O error, dev fd0, sector 0
> Probing devices to guess BIOS drives.  This may take a long time.
> end_requesst: I/O error, dev fd0, sector 0
>
> Please press Enter to activate this console.
>
> It feels like SystemImager is not recreating the partions identical to 
> what they used to be and more importantly, it would seem that it's 
> possible to have a valid partition table (in this case one from rhel3) 
> on a system that you want to replace with a different one (generated 
> from rhel4) can can't.
>
> In summary, I now have 2 images, one for rhel3 and one for rhel4.  The 
> rhel3 WILL install on a system that previously had rhel3 but the rhel4 
> won't install on that same system!  I suppose the next test is to try 
> installing the rhel3 system on one that used to be home for rhel4 but 
> wouldn't you know it, I'm having network problems right now and can't 
> do it.
>
> In any event, I have seen issues in the past with version 3.2.0 where 
> SI got confused when existing partitions weren't consistent with what 
> was trying to be installed and I think this time I have the ideal 
> environment to debug it if you're game.  Just tell me what files you'd 
> like to see and what tests to try.
>
> -mark
>


Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Sisuite-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sisuite-devel

Reply via email to