Could someone help me, I am having trouble restoring a full system
backup tape which contained a good "bru" archive of my full and
updated
Operating system from / down. I had Red Hat Enterprise V3 installed
and about 3 months of updates, which included several updates to the
kernel
that I had noticed the kupdate program downloading and installing. My
latest
kernel on the system was vmlinuz-2.4.21-20.EL. My /etc/grub.conf
stated the
following after the updates.

kernel = /vmlinuz-2.4.21-20.EL Root=LABEL=/
initrd /inird-2.4.21-20.EL.img

kernel = /vmlinuz-2.4.21-15.0.4.EL ro Root=LABEL/
initrd /initrd-2.4.21-15.0.4.EL.img

kernel = /mvlinuz-2.4.21-15.EL ro Root=LABEL/
initrd /initrd-2.4.21-15.EL.img

After I restore the full system backup tape and reboot, the
system
just boots to a black screen with the work "GRUB" and a flashing
cursor.
The system is totally unresponsive to any keyboard keys. I am
guessing
after restoring the tape backup the GRUB program can't locate the now
restored boot image that probably is in a new area on the HD.
I have played around with running the command #grubby
--update-kernel=/boot/vmlinuz-2.4.21-20-.EL to see if it would update
any/or all
necessary entries to find the kernel on boot, but again after the
fifth
time of reloading everything and rebooting the system just comes back
to
a black screen with the word "GRUB".


Any help to this problem would be
greatly appreciated.

Please feel free to email me at "laporte1@charter.net"


Thanks

Dave LaPorte.

Dave wrote:
> Could someone help me, I am having trouble restoring a full system
> backup tape which contained a good "bru" archive of my full and updated
> Operating system from / down. I had Red Hat Enterprise V3 installed
> and about 3 months of updates, which included several updates to the
> kernel that I had noticed the kupdate program downloading and
> installing. My latest kernel on the system was vmlinuz-2.4.21-20.EL.
> My /etc/grub.conf stated the following after the updates.

It needs a little bit more (see below).
>
> kernel = /vmlinuz-2.4.21-20.EL Root=LABEL=/ initrd
> /inird-2.4.21-20.EL.img
>
> kernel = /vmlinuz-2.4.21-15.0.4.EL ro Root=LABEL/ initrd
> /initrd-2.4.21-15.0.4.EL.img
>
> kernel = /mvlinuz-2.4.21-15.EL ro Root=LABEL/ initrd
> /initrd-2.4.21-15.EL.img
>
> After I restore the full system backup tape and reboot, the system just
> boots to a black screen with the work "GRUB" and a flashing cursor. The
> system is totally unresponsive to any keyboard keys. I am guessing
> after restoring the tape backup the GRUB program can't locate the now
> restored boot image that probably is in a new area on the HD. I have
> played around with running the command #grubby
> --update-kernel=/boot/vmlinuz-2.4.21-20-.EL to see if it would update
> any/or all necessary entries to find the kernel on boot, but again
> after the fifth time of reloading everything and rebooting the system
> just comes back to a black screen with the word "GRUB".
>
>
> Any help to this problem would be greatly appreciated.
>
> Please feel free to email me at "laporte1@charter.net"
>
>
> Thanks
>
> Dave LaPorte.

Did you actually use CRU? It is supposed to restore absolutely everything,
partition table, /dev/, and all, by using two floppies and the backup tape.

Is /boot and /boot/grub correct? Sounds like grup is being called and it
cannot find the stuff in /boot/grub.

My /boot/grub looks like this:

]$ ls -l /boot/grub
total 187
-rw-r--r-- 1 root root 82 Mar 2 2004 device.map
-rw-r--r-- 1 root root 7840 Mar 2 2004 e2fs_stage1_5
-rw-r--r-- 1 root root 7536 Mar 2 2004 fat_stage1_5
-rw-r--r-- 1 root root 6880 Mar 2 2004 ffs_stage1_5
-rw------- 1 root root 881 Sep 8 00:42 grub.conf
-rw------- 1 root root 905 Aug 4 01:22 grub.conf.dist
-rw------- 1 root root 905 Aug 4 01:12 grub.conf~
-rw-r--r-- 1 root root 8448 Mar 2 2004 jfs_stage1_5
lrwxrwxrwx 1 root root 11 Mar 2 2004 menu.lst ->
../grub.conf
-rw-r--r-- 1 root root 7040 Mar 2 2004 minix_stage1_5
-rw-r--r-- 1 root root 9408 Mar 2 2004 reiserfs_stage1_5
-rw-r--r-- 1 root root 11182 Sep 11 2003 splash.xpm.gz
-rw-r--r-- 1 root root 512 Mar 2 2004 stage1
-rw-r--r-- 1 root root 106364 Mar 2 2004 stage2
-rw-r--r-- 1 root root 6528 Mar 2 2004 vstafs_stage1_5
-rw-r--r-- 1 root root 9320 Mar 2 2004 xfs_stage1_5


trillian:root[/boot/grub]# cat grub.conf
# grub.conf generated by anaconda
#
# Note that you do not have to rerun grub after making changes to this file
# NOTICE: You have a /boot partition. This means that
# all kernel and initrd paths are relative to /boot/, eg.
# root (hd0,0)
# kernel /vmlinuz-version ro root=/dev/hda3
# initrd /initrd-version.img
#boot=/dev/hda
default=0
timeout=15
splashimage=(hd0,0)/grub/splash.xpm.gz
title Red Hat Enterprise Linux ES (2.4.21-20.ELsmp)
root (hd0,0)
kernel /vmlinuz-2.4.21-20.ELsmp ro root=LABEL=/
initrd /initrd-2.4.21-20.ELsmp.img
title Red Hat Enterprise Linux ES (2.4.21-20.EL)
root (hd0,0)
kernel /vmlinuz-2.4.21-20.EL ro root=LABEL=/
initrd /initrd-2.4.21-20.EL.img
title Red Hat Enterprise Linux ES (2.4.21-15.0.4.ELsmp)
root (hd0,0)
kernel /vmlinuz-2.4.21-15.0.4.ELsmp ro root=LABEL=/
initrd /initrd-2.4.21-15.0.4.ELsmp.img


--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ Registered Machine 241939.
/( )\ Shrewsbury, New Jersey http://counter.li.org
^^-^^ 13:00:00 up 1 day, 23:43, 3 users, load average: 4.36, 4.27, 4.20

Jean-David Beyer <jdbeyer@exit109.com> wrote in message news:<10k3nrqdcjlrg84@corp.supernews.com>...
> Dave wrote:
> > Could someone help me, I am having trouble restoring a full system
> > backup tape which contained a good "bru" archive of my full and updated
> > Operating system from / down. I had Red Hat Enterprise V3 installed
> > and about 3 months of updates, which included several updates to the
> > kernel that I had noticed the kupdate program downloading and
> > installing. My latest kernel on the system was vmlinuz-2.4.21-20.EL.
> > My /etc/grub.conf stated the following after the updates.
>
> It needs a little bit more (see below).
> >
> > kernel = /vmlinuz-2.4.21-20.EL Root=LABEL=/ initrd
> > /inird-2.4.21-20.EL.img
> >
> > kernel = /vmlinuz-2.4.21-15.0.4.EL ro Root=LABEL/ initrd
> > /initrd-2.4.21-15.0.4.EL.img
> >
> > kernel = /mvlinuz-2.4.21-15.EL ro Root=LABEL/ initrd
> > /initrd-2.4.21-15.EL.img
> >
> > After I restore the full system backup tape and reboot, the system just
> > boots to a black screen with the work "GRUB" and a flashing cursor. The
> > system is totally unresponsive to any keyboard keys. I am guessing
> > after restoring the tape backup the GRUB program can't locate the now
> > restored boot image that probably is in a new area on the HD. I have
> > played around with running the command #grubby
> > --update-kernel=/boot/vmlinuz-2.4.21-20-.EL to see if it would update
> > any/or all necessary entries to find the kernel on boot, but again
> > after the fifth time of reloading everything and rebooting the system
> > just comes back to a black screen with the word "GRUB".
> >
> >
> > Any help to this problem would be greatly appreciated.
> >
> > Please feel free to email me at "laporte1@charter.net"
> >
> >
> > Thanks
> >
> > Dave LaPorte.
>
> Did you actually use CRU? It is supposed to restore absolutely everything,
> partition table, /dev/, and all, by using two floppies and the backup tape.
>
> Is /boot and /boot/grub correct? Sounds like grup is being called and it
> cannot find the stuff in /boot/grub.
>
> My /boot/grub looks like this:
>
> ]$ ls -l /boot/grub
> total 187
> -rw-r--r-- 1 root root 82 Mar 2 2004 device.map
> -rw-r--r-- 1 root root 7840 Mar 2 2004 e2fs_stage1_5
> -rw-r--r-- 1 root root 7536 Mar 2 2004 fat_stage1_5
> -rw-r--r-- 1 root root 6880 Mar 2 2004 ffs_stage1_5
> -rw------- 1 root root 881 Sep 8 00:42 grub.conf
> -rw------- 1 root root 905 Aug 4 01:22 grub.conf.dist
> -rw------- 1 root root 905 Aug 4 01:12 grub.conf~
> -rw-r--r-- 1 root root 8448 Mar 2 2004 jfs_stage1_5
> lrwxrwxrwx 1 root root 11 Mar 2 2004 menu.lst ->
> ./grub.conf
> -rw-r--r-- 1 root root 7040 Mar 2 2004 minix_stage1_5
> -rw-r--r-- 1 root root 9408 Mar 2 2004 reiserfs_stage1_5
> -rw-r--r-- 1 root root 11182 Sep 11 2003 splash.xpm.gz
> -rw-r--r-- 1 root root 512 Mar 2 2004 stage1
> -rw-r--r-- 1 root root 106364 Mar 2 2004 stage2
> -rw-r--r-- 1 root root 6528 Mar 2 2004 vstafs_stage1_5
> -rw-r--r-- 1 root root 9320 Mar 2 2004 xfs_stage1_5
>
>
> trillian:root[/boot/grub]# cat grub.conf
> # grub.conf generated by anaconda
> #
> # Note that you do not have to rerun grub after making changes to this file
> # NOTICE: You have a /boot partition. This means that
> # all kernel and initrd paths are relative to /boot/, eg.
> # root (hd0,0)
> # kernel /vmlinuz-version ro root=/dev/hda3
> # initrd /initrd-version.img
> #boot=/dev/hda
> default=0
> timeout=15
> splashimage=(hd0,0)/grub/splash.xpm.gz
> title Red Hat Enterprise Linux ES (2.4.21-20.ELsmp)
> root (hd0,0)
> kernel /vmlinuz-2.4.21-20.ELsmp ro root=LABEL=/
> initrd /initrd-2.4.21-20.ELsmp.img
> title Red Hat Enterprise Linux ES (2.4.21-20.EL)
> root (hd0,0)
> kernel /vmlinuz-2.4.21-20.EL ro root=LABEL=/
> initrd /initrd-2.4.21-20.EL.img
> title Red Hat Enterprise Linux ES (2.4.21-15.0.4.ELsmp)
> root (hd0,0)
> kernel /vmlinuz-2.4.21-15.0.4.ELsmp ro root=LABEL=/
> initrd /initrd-2.4.21-15.0.4.ELsmp.img


************************************************** *********
Jean-David Beyer:

I Didn't use CRU I used a tape software package called "BRU"
which
is supposed to back everything during a full system backup. When I
restore
I "telinit s" to single user mode and then I restore using the brue
command #bru -xvua. The -ua means overwrite all files. I have used
this process
on older redhat versions and it worked. The only other thing I did to
insure
the system booted after the tape restore was to run the 'lilo' command
to
reset the kernel info.
In regard to the /boot and /boot/grub the only thing I have been
checking
is the /boot directory to make sure the the file vmlinuz-2.4.21-20.EL
and the
file initrd-2.4.21-20.ELsmp.img were actually there and they were. I
didn't
even know to check the /boot/grub. I will try again and compare the
files to
what you posted.
Also, I did check the /etc/grub.conf. It was there and it
looked similar to what you had other then my vmlinux files were
numbered

vmlinuz-2.4.21-20.EL
vmlinuz-2.4.21-15.0.4.EL
vmlinuz-2.4.21-15.0.4.EL

Before I restored the system the only entry in /etc/grub.conf was
for the kernel build vmlinuz-2.4.21-15.0.4.EL. I assumed that was
because my tape backup was of a system with several Red Hat updates.

Thanks


Dave.

Dave wrote:

> Jean-David Beyer:
>
> I Didn't use CRU I used a tape software package called "BRU"
> which
> is supposed to back everything during a full system backup.

I know BRU and use it myself. It will backup all the files you specify.
CRU is a free package (that requires BRU) that will enable a bare-metal
restore. Not only does it restore files, it uses information (at backup
time) from fdisk so that it can remake the partition tables, etc. I.e., if
your machine gets hit by lightning, all you need do is buy a new one with
a suitable tape drive, and disk drives at least as large as the old ones,
put in the two floppies and the tape and be back where you started. It
remakes the partitions, sets up /dev with all your permissions, etc.

> When I
> restore
> I "telinit s" to single user mode and then I restore using the brue
> command #bru -xvua. The -ua means overwrite all files. I have used
> this process
> on older redhat versions and it worked. The only other thing I did to
> insure
> the system booted after the tape restore was to run the 'lilo' command
> to
> reset the kernel info.

Assuming your partition table was not messed up, and you really backed up
everything under / (root), I would expect this to have worked. Assuming
you have a separate /boot partition properly located, or a machine where
the 1024 cylinder limit is not a problem (newer machines do not have this
problem).

> In regard to the /boot and /boot/grub the only thing I have been
> checking
> is the /boot directory to make sure the the file vmlinuz-2.4.21-20.EL
> and the
> file initrd-2.4.21-20.ELsmp.img were actually there and they were. I
> didn't
> even know to check the /boot/grub. I will try again and compare the
> files to
> what you posted.
> Also, I did check the /etc/grub.conf. It was there and it
> looked similar to what you had other then my vmlinux files were
> numbered
>
> vmlinuz-2.4.21-20.EL
> vmlinuz-2.4.21-15.0.4.EL
> vmlinuz-2.4.21-15.0.4.EL
>
> Before I restored the system the only entry in /etc/grub.conf was
> for the kernel build vmlinuz-2.4.21-15.0.4.EL. I assumed that was
> because my tape backup was of a system with several Red Hat updates.

In other words, you installed several updates without making a new backup?
In that case, you better have vmlinuz-2.4.21-15.0.4.EL ohn the machine as
well as it may try to boot that. But from your original description, you
do not even get to the point where it asks which version you want. Perhaps
grub does not ask if you have only one. Since I never have less than two
(SMP and UP), I do not know what it does.

Do you have splash.xpm.gz in /boot/grub?
>
> Thanks
>
>
> Dave.


--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ Registered Machine 241939.
/( )\ Shrewsbury, New Jersey http://counter.li.org
^^-^^ 08:15:00 up 2 days, 18:58, 3 users, load average: 4.17, 4.19, 4.14

Jean-David Beyer <jdbeyer@exit109.com> wrote in message news:<10k5rrigge5vocf@corp.supernews.com>...
> Dave wrote:
>
> > Jean-David Beyer:
> >
> > I Didn't use CRU I used a tape software package called "BRU"
> > which
> > is supposed to back everything during a full system backup.
>
> I know BRU and use it myself. It will backup all the files you specify.
> CRU is a free package (that requires BRU) that will enable a bare-metal
> restore. Not only does it restore files, it uses information (at backup
> time) from fdisk so that it can remake the partition tables, etc. I.e., if
> your machine gets hit by lightning, all you need do is buy a new one with
> a suitable tape drive, and disk drives at least as large as the old ones,
> put in the two floppies and the tape and be back where you started. It
> remakes the partitions, sets up /dev with all your permissions, etc.
>
> > When I
> > restore
> > I "telinit s" to single user mode and then I restore using the brue
> > command #bru -xvua. The -ua means overwrite all files. I have used
> > this process
> > on older redhat versions and it worked. The only other thing I did to
> > insure
> > the system booted after the tape restore was to run the 'lilo' command
> > to
> > reset the kernel info.
>
> Assuming your partition table was not messed up, and you really backed up
> everything under / (root), I would expect this to have worked. Assuming
> you have a separate /boot partition properly located, or a machine where
> the 1024 cylinder limit is not a problem (newer machines do not have this
> problem).
>
> > In regard to the /boot and /boot/grub the only thing I have been
> > checking
> > is the /boot directory to make sure the the file vmlinuz-2.4.21-20.EL
> > and the
> > file initrd-2.4.21-20.ELsmp.img were actually there and they were. I
> > didn't
> > even know to check the /boot/grub. I will try again and compare the
> > files to
> > what you posted.
> > Also, I did check the /etc/grub.conf. It was there and it
> > looked similar to what you had other then my vmlinux files were
> > numbered
> >
> > vmlinuz-2.4.21-20.EL
> > vmlinuz-2.4.21-15.0.4.EL
> > vmlinuz-2.4.21-15.0.4.EL
> >
> > Before I restored the system the only entry in /etc/grub.conf was
> > for the kernel build vmlinuz-2.4.21-15.0.4.EL. I assumed that was
> > because my tape backup was of a system with several Red Hat updates.
>
> In other words, you installed several updates without making a new backup?
> In that case, you better have vmlinuz-2.4.21-15.0.4.EL ohn the machine as
> well as it may try to boot that. But from your original description, you
> do not even get to the point where it asks which version you want. Perhaps
> grub does not ask if you have only one. Since I never have less than two
> (SMP and UP), I do not know what it does.
>
> Do you have splash.xpm.gz in /boot/grub?
> >
> > Thanks
> >
> >
> > Dave.


Jean-David Beyer:

I have to apologize I didn't realize that CRU was associated with
BRU. I will definitely look into using CRU vs doing a complete reinstall
of an os when restoring a full system back.
Also, now after reading your last post I'm seriously questioning my
backups integrity. I viewed the brulog that I created when doing my daily
backup but I wonder if my /root dir really was backed up. I'm going to
dump my tape's contents to a file and view it.

Question: When you backup your Red Hat Ent v3 system do you just
run it off a cron scripts and use this Bru command,

/bin/bru -cv / >> /etc/brulog 2>&1

That is typically how I run my full system backups. I don't create that
backup from single user mode.

Dave.

Dave wrote:

> I have to apologize I didn't realize that CRU was associated with
> BRU. I will definitely look into using CRU vs doing a complete reinstall
> of an os when restoring a full system back.
> Also, now after reading your last post I'm seriously questioning my
> backups integrity. I viewed the brulog that I created when doing my daily
> backup but I wonder if my /root dir really was backed up. I'm going to
> dump my tape's contents to a file and view it.
>
> Question: When you backup your Red Hat Ent v3 system do you just
> run it off a cron scripts and use this Bru command,
>
> /bin/bru -cv / >> /etc/brulog 2>&1
>
> That is typically how I run my full system backups. I don't create that
> backup from single user mode.
>
My /etc/cron.daily contains an entry:

-rwxr-xr-- 1 root root 2294 May 6 06:38 zBackup.cron

That file is shown below. It may do more than you want, but it has some
conveniences. For example, it is easy to stop the cron job while I am on
vacation and not piddle with anything else. My /etc/crontab runs
/etc/cron.daily only Monday-Saturday, and /etc/cron.weekly only on
Sundays. There are mostly links from /etc/cron.weekly to /etc/cron.daily,
but the Sunday backup is different from the weekly one in case I wish to
do incremental daily backups during the week and full ones weekly. Since
right now everything fits onto one tape, it is simpler to do a full backup
every time.

VXA-tool is a tool that will read out stuff in the tape drive when an
error occurs. Exabyte provide this free. You do not absolutely need it,
but if you get a problem, it is good to have its output.


#!/bin/bash

# Weekday backup to tape.

# Global definitions.

declare -i EXIT_STATUS ERROR GOOD WARNING EXIT_STATUS_MT

BACKUPDIR=/var/log/TapeBackups
ERROR=2
GOOD=0
TAPE_DRIVE=/dev/st0
WARNING=1

BRU=/bin/bru
CHGRP=/bin/chgrp
CHMOD=/bin/chmod
CPIO=/bin/cpio
ECHO=/bin/echo
FIND=/usr/bin/find
MAIL=/bin/mail
ME=jdbeyer
MT=/bin/mt
MV=/bin/mv
TAIL=/usr/bin/tail
VXATOOL=/usr/local/Downloaded/ECRIX/vxaTool

REPORT=$BACKUPDIR/report.`/bin/date +%A`

# Has someone asked us to skip backups?
if [ -f $BACKUPDIR/skip.backups ] ; then
$ECHO `/bin/date` "Backup skipped by request." >> $BACKUPDIR/backups
exit 0
fi

$ECHO `/bin/date` "Doing Weekday backup." >> $BACKUPDIR/backups
$TAIL --lines 100 --quiet $BACKUPDIR/backups >/tmp/backups
$MV /tmp/backups $BACKUPDIR/backups
$CHGRP db2iadm1 $BACKUPDIR/backups
$CHMOD 0664 $BACKUPDIR/backups

# Do the weekday backup.
$ECHO `/bin/date` "Weekday backup for trillian." > $REPORT
$BRU -c -L "`/bin/date +%A%W`" -V -B -X / >> $REPORT 2>&1
EXIT_STATUS=$?
case $EXIT_STATUS in
$ERROR )
$ECHO "bru failed with Error: logfile attempted" >> $REPORT
$VXATOOL $TAPE_DRIVE -d $BACKUPDIR/logfile`/bin/date +%Y%m%d%H%M`.bin
$ECHO "bru create failed with Error" >> $REPORT
$CHMOD 0664 $REPORT
$MT -f $TAPE_DRIVE offline
$MAIL -s "bru Report" $ME <$REPORT
exit 1
;;
$WARNING )
$ECHO "bru returned Warning" >> $REPORT
$CHMOD 0664 $REPORT
$MT -f $TAPE_DRIVE offline
$MAIL -s "bru Report" $ME <$REPORT
exit 1
;;

$GOOD )
$CHMOD 0664 $REPORT
$MT -f $TAPE_DRIVE offline
EXIT_STATUS_MT=$?
case $EXIT_STATUS_MT in
$ERROR )
$ECHO "mt offline failed with Error: logfile attempted" >> $REPORT
$VXATOOL $TAPE_DRIVE -d $BACKUPDIR/logfile`/bin/date +%Y%m%d%H%M`.bin
$ECHO "mt offline failed with Error" >> $REPORT
;;
$WARNING )
$ECHO "mt offline failed with Error: invalid device" >> $REPORT
;;
$GOOD )
;;
* )
$ECHO "mt returned unexpected: " $EXIT_STATUS_MT >> $REPORT
;;
esac
$MAIL -s "bru Report" $ME <$REPORT
exit 0
;;

* )
$ECHO "bru returned unexpected: " $EXIT_STATUS >> $REPORT
$CHMOD 0664 $REPORT
$MT -f $TAPE_DRIVE offline
$MAIL -s "bru Report" $ME <$REPORT
exit 1
;;
esac


--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ Registered Machine 241939.
/( )\ Shrewsbury, New Jersey http://counter.li.org
^^-^^ 11:20:00 up 4 days, 22:03, 5 users, load average: 2.28, 2.21, 2.17