system backup - Backup fails

system backup - Backup fails

Post by Silverstei » Fri, 04 Dec 1998 04:00:00



I have an irix 6.2 system with an xfs disk.  I wanted to do a full
system backup.  According to the insight guide one should use the Backup
program (which is a front-end for the bru program).  So I ran

"bru /"

This chugs along, then asks to insert a second 4mm DAT which I did, and
then gives:

.
.
.
c      2k of 1110796k [2] /lib32
c    746k of 1111542k [2] /lib32/rld
c   2330k of 1113872k [2] /lib32/libc.so.1
c      4k of 1113876k [2] /.login
c      2k of 1113878k [2] /floppy
c      2k of 1113880k [2] /.swap.virtual
c      4k of 1113884k [2] /.profile
c      2k of 1113886k [2] /._mmod
/usr/sbin/Backup failed (exit status 1), /etc/lastbackup not updated

Backup complete.
Backup interrupted

So why?

What I really want to know is what is the best way to do a complete
system backup?   Which program should I use (In case the disk gets
trashed I'd like to be able to as trivially as possible just load
everything it back - os, programs, user files, etc. -  from a 4mm tape).

Also, which programs can be used from the command monitor?

Herc

 
 
 

system backup - Backup fails

Post by scott sisso » Fri, 04 Dec 1998 04:00:00



> I have an irix 6.2 system with an xfs disk.

Try using xfsdump for xfs filesystems...

    xfsdump -o -f /dev/tape -l 0 /

The -l 0 is dump level zero.  Incrementals
would required -l 1, -l 2 etc...

xfsdump -I will produce a history report
from data stored in /var/xfsdump/inventory
when each dump is completed. xfsdumps can
be interrupted and resumed.

man xfsdump

Warning: if the system crashes, some older 6.2
CDs if I recall do not have the xfsdump/xfsrestore
combination in mini-root but an SGI tech would
know for sure.

Kind Regards,
scott sisson, consultant
Numerical Integrations, Inc.

http://www.numerang.com

 
 
 

system backup - Backup fails

Post by Ralf Beye » Sat, 05 Dec 1998 04:00:00


        I have an irix 6.2 system with an xfs disk.  I wanted to do a full
        system backup.  According to the insight guide one should use the Backup
        program (which is a front-end for the bru program).  So I ran

        "bru /"

        This chugs along, then asks to insert a second 4mm DAT which I did, and
        then gives:

        .
        .
        .
        c      2k of 1110796k [2] /lib32
        c    746k of 1111542k [2] /lib32/rld
        c   2330k of 1113872k [2] /lib32/libc.so.1
        c      4k of 1113876k [2] /.login
        c      2k of 1113878k [2] /floppy
        c      2k of 1113880k [2] /.swap.virtual
        c      4k of 1113884k [2] /.profile
        c      2k of 1113886k [2] /._mmod
        /usr/sbin/Backup failed (exit status 1), /etc/lastbackup not updated

        Backup complete.
        Backup interrupted

If this happens, the first line reflects not
exactly what has been done and the second line
is wrong.

This may happen if some files were not backed
up like swap files or directories/files with
path names too long. Use the -v option of bru
to see which directories/files were not backed
up.

Interpret 'Backup complete' as 'Backup complete
as far as possible'.

Ignore the second line or change line 140 in
/usr/sbin/Backup from:
  trap 'cd /; echo Backup interrupted; rm -rf $tabfiles; exit 1' 0 2
to:
  trap 'cd /; echo Backup interrupted; rm -rf $tabfiles; exit 1' 2

        What I really want to know is what is the best way to do a complete
        system backup?   Which program should I use (In case the disk gets
        trashed I'd like to be able to as trivially as possible just load
        everything it back - os, programs, user files, etc. -  from a 4mm tape).

        Also, which programs can be used from the command monitor?

Why do you not use 'Backup' from the Toolchest
or /usr/sbin/Backup from the command moonitor
and in case of a crash SGI's Crash Recovery
procedure ?

Regards


German Aerospace Center (DLR)
Deutsches Zentrum fuer Luft- und Raumfahrt e.V.

 
 
 

1. Backup and Restore Manager skips root disk during Full Backup

I replaced my system (root) disk on my Power Indigo2 (IRIX 6.5.17m)
last week via a disk cloning procedure.  After formatting the drive
with fx I copied over the volume header files via dvhtool and used
xfsdump/xfsrestore to copy the files.  Everything went smoothly and
the new disk works just fine.  The filesystem looks good and I have no
problems accessing or creating any files on it.

Unfortunately, when I started a Full backup using the Backup and
Restore Manager, the root disk was apparently skipped.  All other
internal and external SCSI disks were properly backed up.  I tried
starting the procedure again this morning using the
/usr/sysadm/privbin/backup command instead of the GUI process, but I'm
getting similar results.

This is the second time I've replaced the system disk via cloning (did
it three years ago) and this did not occur the first time.  The only
thing I can think of that may be confusing Backup is that the new disk
is quite large (50 Gb).  Is there a size limit to filesystems that
Backup will try to process?

Thanks in advance,
Alan Glendinning

2. CP/M software

3. Backup doesn't backup swap file

4. Fantastic news for all *REAL* OS/2 users

5. Backup and Restore Manager skips root disk during Full Backup

6. Does my ISDN router need the ISP assigned static IP?

7. How does the Backup GUI communicate with the backup process

8. ASCII picture editor?

9. remote tape backup failed - but filled sysdisk...

10. Backup failed?

11. Indigo backup battery clock failed

12. Mixed environment network backup system

13. WHole BAckup of the entire system