Topic: Unable to boot after installing Update2??

[Jason Evans]

I installed the update2 on my SME 5.5 server. Everything seemed to go ok. However just after POST the Server could'nt find the boot record. Its just whent LLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLL

I'm running the server on a SCSI hard disk, which is working ok.
So I tried to boot from the installation repair disk and it said it could'nt find the kernel image. So I typed esmith and it booted much further but just before it would display your hostname Kernel Panic. I booted from CD and ran the upgrade,and just as it was finishing the upgrade it gave me an error saying that the boot files were not on the scsi drive, and gave me an error that the Server is unstable. I rebooted the server again and was still unable to boot from the hard disk. I booted from the floppy, and again it said it could'nt find the image. However I was able to boot from the esmith image on the floppy this time.

So the server is up and running now, but I need the floppy in the drive to boot from the e-smith image.

So what is the best way to repair this server? Do I need to restore the mbr and copy the Linux image back to disk?

Hope someone can help, don't have a clue how this happened.

[Jason Evans]

After thinking about this a little more I don't think it has anything to do with installing the Update2. I just haven't had a need to reboot until I updated the RPM's with Update2. I remember when I first installed SME 5.5 the configuration file size in the backup and restore web page was over 200MB I beleive. Now just recently it says "Currently your configuration and data files total approximately 15Mb" I beleive I'm missing the Linux kernel image on the hard drive and countless other config files.

At any rate does anyone know how to repair this problem. I don't want to reinstall the Server since its currently in production, and its backup brothers Motherboard coughed up a chip.

[Jason Evans]

After thinking about this a little more I don't think it has anything to do with installing the Update2. I just haven't had a need to reboot until I updated the RPM's with Update2. I remember when I first installed SME 5.5 the configuration file size in the backup and restore web page was over 200MB I beleive. Now just recently it says "Currently your configuration and data files total approximately 15Mb" I beleive I'm missing the Linux kernel image on the hard drive and countless other config files.

At any rate does anyone know how to repair this problem. I don't want to reinstall the Server since its currently in production, and its backup brothers Motherboard coughed up a chip.

[NickR]

I'd suspect that something fsck'd your  /boot  partion.

First thing I'd do is run fdisk & look at the status of the partions - I suspect that you'll find the partion type is wrongly set.

[Jason Evans]

Thanks for helping Nick,

Well, here's what the partition tables look like.

My first SCSI boot disk.

Disk /dev/sda: 255 heads, 63 sectors, 522 cylinders
Units = cylinders of 16065 * 512 bytes

   Device Boot    Start       End    Blocks   Id  System
/dev/sda1   *         1         2     16033+  83  Linux
/dev/sda2             3       522   4176900    5  Extended
/dev/sda5             3        35    265041   82  Linux swap
/dev/sda6            36       522   3911796   83  Linux

My IDE disk.

Disk /dev/hda: 255 heads, 63 sectors, 4866 cylinders
Units = cylinders of 16065 * 512 bytes

   Device Boot    Start       End    Blocks   Id  System
/dev/hda1             1      4866  39086113+  83  Linux

Is there a problem with the start and end blocks??

[NickR]

Hmm, that's certainly different to mine (OK, mine is a RAID1 server, but still...)

Disk /dev/sda: 255 heads, 63 sectors, 2232 cylinders
Units = cylinders of 16065 * 512 bytes
 
   Device Boot    Start       End    Blocks   Id  System
/dev/sda1   *         1        33    265041   fd  Linux raid autodetect
/dev/sda2            34      2232  17663467+   5  Extended
/dev/sda5            34        35     16033+  fd  Linux raid autodetect
/dev/sda6            36      2232  17647371   fd  Linux raid autodetect
 
Disk /dev/sdb: 255 heads, 63 sectors, 2232 cylinders
Units = cylinders of 16065 * 512 bytes
 
   Device Boot    Start       End    Blocks   Id  System
/dev/sdb1   *         1        33    265041   fd  Linux raid autodetect
/dev/sdb2            34      2232  17663467+   5  Extended
/dev/sdb5            34        35     16033+  fd  Linux raid autodetect
/dev/sdb6            36      2232  17647371   fd  Linux raid autodetect


I think the clue is in the number of blocks on /dev/sda1 and /dev/sda5 - you'll notice that these are reversed between yours & mine.  I don't know how you'd recover from this, but it looks to me as though it's swapped the /boot & the swap partions, so you're now trying to boot from the swap partition, which is obviously not good!

I think I'd be inclined to take a backup to desktop & re-install clean, unless someone else can suggest a better option.

[Charlie Brady]

NickR wrote:

> Hmm, that's certainly different to mine (OK, mine is a RAID1
> server, but still...)

So who knows how it could be expected to be different.

> I think the clue is in the number of blocks on /dev/sda1 and
> /dev/sda5 - you'll notice that these are reversed between
> yours & mine.  I don't know how you'd recover from this, but
> it looks to me as though it's swapped the /boot & the swap
> partions, so you're now trying to boot from the swap
> partition, which is obviously not good!

The update doesn't touch the partition structure on disk, so this isn't at all likely to be relevant.

> I think I'd be inclined to take a backup to desktop &
> re-install clean, unless someone else can suggest a better
> option.

The first things that Jason should do is look at lilo documentation to understand the LLL... issue. Once he's booted from floppy, he could  run "/sbin/lilo -v" and look for any error output. Re-running lilo is one of the things that "signal-event post-upgrade" does.

Charlie

[Jason Evans]

Thanks for you input guys!

I am getting very frustrated now. I have since reinstalled the server. Actually I have reinstalled it a whole bunch of times. I am convinced that the update2 does something the boot records or even / root.

This is whats happening......

I install fresh onto 1 scsi disk with no partitions.

I then boot the server and configure it from the console with the standart settings IP address DNS ect...

I then shutdown and install a second scsi disk /dev/sdb1 and a IDE disk /dev/hda1.

I then add a few options from the server-manager. User accounts, groups and 2 ibays. I then test mount the drives to the ibays OK.

I then add the drives to the /etc/fstab /dev/sdb1 /home/e-smith/files/ibays/ibayname
/dev/hda1/home/e-smith/files/ibays/ibayname

I then reboot and everything is OK, drives mounted.

I install Update1 from the server-manager.

Reboot everything ok.

Now heres were the problem start.

I install update2 from the RPMS directory on the hard drive that I downloaded from e-smith. Using this syntax update the rpms.

rpm -Uvh --replacepkgs *.rpm
 
Then
 
/sbin/e-smith/signal-event post-upgrade
 
Then
 
/sbin/e-smith/signal-event reboot.

When the server reboots I get this LI.

If I boot from the floppy I get Kernel panic unable to mount / filesystem.

Whats going on here guys!!!!!!!!!!!!

[Andy Berry]

Similar problems for me, but I installed the Update from The Blades page in Server manager.  
I got tons of errors upon reboot, fsck indicated problems with several blocks and duplicate data.  It came up after running the check, but died again after deleting a share.
I'm just dowloading the latest ISO to reinstall. Fortunately this machine isn't in production yet.
-A

[Jason Evans]

Well I have found a way to reproduce this disasterous problem that I have been running into. Me being a newbee I can only tell you what I've observed rather than to explain why its happening. I would call it a bug, you guru's can tell me different. I'm probably doing something stupid along the way.

This is what I have found....

Senario 1

Install SME 5.5 on a scsi drive.

Configure the server with mandatory settings.

Install Update1 from server-manager.

Reboot

Install Update2 from RPMS.

Reboot

All is good!!!!

Add a second hard drive and add it to fstab.

All is good.


However if you do it like the following, the outcome is a little different..........

Senario 2

Install SME 5.5 on a scsi drive.

Configure the server with mandatory settings.

Install Update1 from server-manager.

Install a second hard disk, and add it to fstab.

Install Update2, after reboot the server is pooched??

If you remove the second drive before you install Update2 it works fine.

I wanted to install the second and third drives after SME was installed since I found that when I had 2 scsi drives ID's 0 and 1 in the system at the same time /boot would be installed on scsi 0 and / would be installed on scsi1.

I did read up on LILO boot errors which was suggested since I was getting L LI or LLLLLLLLLLLLLLLLLLL.

L

The first stage boot loader has been loaded and started, but it can't load the second stage boot loader. The two-digit error codes indicate the type of problem. (See also section "Disk error codes".) This condition usually indicates a media failure or a geometry mismatch (e.g. bad disk parameters, see section "Disk geometry").


 LI

The first stage boot loader was able to load the second stage boot loader, but has failed to execute it. This can either be caused by a geometry mismatch or by moving /boot/boot.b without running the map installer.


LIL

LIL The second stage boot loader has been started, but it can't load the descriptor table from the map file. This is typically caused by a media failure or by a geometry mismatch.The second stage boot loader has been loaded at an incorrect address. This is typically caused by a subtle geometry mismatch or by moving /boot/boot.b without running the map installer.

 LIL- The descriptor table is corrupt. This can either be caused by a geometry mismatch or by moving /boot/map without running the map installer.


 LILO All parts of LILO have been successfully loaded.

*Try these steps if you get a LI and your LBA is enabled in your BIOS already:

Boot with your Red Hat install floppy.
 Answer questions up to the INSTALL/UPGRADE question and pick upgrade.
 Wait for next question to appear.
 Hit -
 mount your root partition: mount /dev/hda1 /mnt
 re-run lilo: /mnt/sbin/lilo -r /mnt
 unmount your root partition: umount /mnt
 reboot.


 Mount

mount is the command that one would need to add/remove hardware from your linux machine.
mount /dev/cdrom /mnt/cdrom
ok mount is to mount
/dev/cdrom is the thing you want mounted
/mnt/cdrom is where linux mounts it so you can see it
the same apply to floppy drives
floppy drives fd0
I use mount /dev/fd0 /mnt/floppy


Thanks for the help guys