oops, I was wrong.. anyway the cylinder boundary alignement is not a
real problem IMHO. It could be only if you need, for some strange
reasons, to use the old CHS geometry with your disk (maybe if you've an
old disk, but in general it can be considered deprecated and totally
replaced by LBA).

You see those warnings because fdisk still uses the old addressing
(CHS), parted instead seems to ignore the cylinder alignment:

http://www.gnu.org/software/parted/manual/html_mono/parted.html#SEC34

Moreover I don't understand why the patch solved your grub problem...
only stage1 must be installed in the sector 0 (MBR), then stage2 blocks
are loaded by stage1 that knows the address of all the stage2 blocks,
but stage2 can be anywhere in the disk, in general not before the 1st
MB... can you post your partitioning schema (autoinstallscript.conf)?

-Andrea

e e wrote:
> The patch did not fix the cylinder boundary problem. It however solved
> issue #2 (grub installation) As we left 1 MB vacant before the first
> partition (root partition) the boot loader had enough space to install
> stage1 and stage 2. Still not sure why grub install is sucessful in
> /dev/hda and why it failed in /dev/hdb. 
> 
> The cylinder boundary problem is still coming.
> 
> I was previously using system imager version 3.2.3 over redhat 9.  I did
> not face the cylinder boundary issue in that version.  
> 
> I tried to use fdisk instead of parted. I tried to include fdisk in boel
> binaries tar.gz, But there is libc incompatibility. Is there a fdisk
> compatible with boel.
> 
> 
> 
> 
> 
> On 8/2/06, *Andrea Righi* <[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>>
> wrote:
> 
>     Regarding the first problem (partition bondaries), it seems that fdisk
>     expects an unused block between two contiguous partitions. With parted
>     this is simply ignored, so I don't think it's a problem, but IMHO the
>     fdisk approach is cleaner, so we should use it...
> 
>     Moreover the patch is very simple:
> 
>     Index: lib/SystemImager/Server.pm
>     ===================================================================
>     --- lib/SystemImager/Server.pm  (revision 3704)
>     +++ lib/SystemImager/Server.pm  (working copy)
>     @@ -694,14 +694,14 @@
> 
>                  # Leave info behind for the next partition. -BEF-
>                  if ("$p_type{$m}" eq "primary") {
>     -                print $out q(END_OF_LAST_PRIMARY=$END_MB) . qq(\n);
>     +                print $out q(END_OF_LAST_PRIMARY=$(($END_MB + 1))) .
>     qq(\n);
> 
>                  } elsif ("$p_type{$m}" eq "extended") {
>     -                print $out q(END_OF_LAST_PRIMARY=$END_MB) . qq(\n);
>     +                print $out q(END_OF_LAST_PRIMARY=$(($END_MB + 1))) .
>     qq(\n);
>                      print $out q(END_OF_LAST_LOGICAL=$START_MB) . qq(\n);
> 
>                  } elsif ("$p_type{$m}" eq "logical") {
>     -                print $out q(END_OF_LAST_LOGICAL=$END_MB) . qq(\n);
>     +                print $out q(END_OF_LAST_LOGICAL=$(($END_MB + 1))) .
>     qq(\n);
>                  }
> 
>                  #
> 
>     Regarding the other problem systemconfigurator always installs grub in
>     the MBR of (hd0). The hd0 disk is mapped by grub-install using the
>     --recheck option and it depends on the first disk it's able to probe.
> 
>     A workaround could be to create a post-install script like the
>     following:
> 
>     -----------------
>     #!/bin/sh
>     grub-install --no-floppy --force-lba /dev/hdb
>     -----------------
> 
>     Since post-install scripts are executed after systemconfigurator you can
>     override its actions (in this case restricted to the bootloader
>     installation).
> 
>     If it works you can implement something more smart to autodetect the
>     opportune disk to use (for example if you have both clients with disk in
>     /dev/hda and others in /dev/hdb).
> 
>     Regards,
>     -Andrea
> 
>     e e wrote:
>     > I am not able to even manually install grub in /dev/hdb. It says
>     no free
>     > sectors available. I feel that both these issues are related. I think
>     > once the parted problem is fixed, boot loader installation will
>     also be
>     > sucessfull. Please correct me if I am wrong.
>     >
>     > In the mean time can you suggest anything to work around the
>     problem???
>     > If there is no workaround possible, I will use sfdisk/fdisk instead of
>     > parted and do the partitioning myself in the autoinstall.template
>     file.
>     > Is it OK???
>     >
>     > Thanks a lot.
>     >
>     >
> 
>     -------------------------------------------------------------------------
>     Take Surveys. Earn Cash. Influence the Future of IT
>     Join SourceForge.net 's Techsay panel and you'll get the chance to
>     share your
>     opinions on IT & business topics through brief surveys -- and earn cash
>     http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>     
> <http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV>
>     _______________________________________________
>     Sisuite-users mailing list
>     [email protected]
>     <mailto:[email protected]>
>     https://lists.sourceforge.net/lists/listinfo/sisuite-users
> 
> 

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Sisuite-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Reply via email to