Koozali.org: home of the SME Server

Backup

Tim Hogan

Backup
« on: June 02, 2002, 02:58:15 AM »
I am running a nightly backup to tape using the backup utility in Server-Manager.  The message sent to admin (see below) looks like it is doing two dumps:  File 1 says it is backing up / and File 2 says it is backing up /boot.  When I follow the directions on the flexbackup-howto.html from myezserver.com to list the contents of the tape it only shows the contents of /boot.  Using the interactive restore features in the same howto I also only see the contents of /boot.  If I try running restore-now to restore a file to an ibay it says the directory is not on the tape.

I have not edited any of the backup or flexbackup configuration files.  Do I need to edit something to have the ibays backed up?  Why does it show / being backed up (File 1) but I don't see it on the tape?  It says that 1542.85MB were backed up.  Is File 2 (/boot) overwriting File 1 (/)?

Thanks,

Tim Hogan

The message I receive in the admin mailbox reads:

flexbackup version 0.9.8
/etc/flexbackup.conf syntax OK

|------------------------------------------------
| Trying "mt defblksize" instead of "mt setblk"
| Doing level 0 backup of all using dump
| Retensioning tape...
| Rewinding & erasing tape...
| Creating index key 200206010203.38
| Tape #0
| Filesystems = / /boot
|------------------------------------------------
At block 0.
|------------------------------------------------
| File number 1, index key 200206010203.38
| Backup of: /
| Date of this level 0 backup: Sat Jun 01 02:04:22 2002
| Date of last level 0 backup: the epoch
|------------------------------------------------
| (dump -0 -b 10 -a -f - / | gzip -4) | buffer -m 3m -s 10k -u 100 -t -p 75 -B \
|  -o /dev/st0
|------------------------------------------------
  DUMP: Date of this level 0 dump: Sat Jun  1 02:04:23 2002
  DUMP: Date of last level 0 dump: the epoch
  DUMP: Dumping /dev/md1 (/) to standard output
  DUMP: Label: none
  DUMP: mapping (Pass I) [regular files]
  DUMP: mapping (Pass II) [directories]
  DUMP: estimated 1516966 tape blocks.
  DUMP: Volume 1 started at: Sat Jun  1 02:08:01 2002
  DUMP: dumping (Pass III) [directories]
  DUMP: dumping (Pass IV) [regular files]
  DUMP: 10.20% done, finished in 0:44
  DUMP: 24.25% done, finished in 0:31
  DUMP: 35.51% done, finished in 0:27
  DUMP: 43.37% done, finished in 0:26
  DUMP: 53.12% done, finished in 0:22
  DUMP: 65.28% done, finished in 0:16
  DUMP: 72.97% done, finished in 0:13
  DUMP: 81.58% done, finished in 0:09
  DUMP: 92.32% done, finished in 0:03
  DUMP: 100.00% done, finished in 0:00
  DUMP: Volume 1 completed at: Sat Jun  1 02:58:54 2002
  DUMP: Volume 1 took 0:50:53
  DUMP: Volume 1 transfer rate: 517 KB/s
  DUMP: 1579883 tape blocks (1542.85MB)
  DUMP: finished in 3053 seconds, throughput 517 KBytes/sec
  DUMP: Date of this level 0 dump: Sat Jun  1 02:04:23 2002
  DUMP: Date this dump completed:  Sat Jun  1 02:58:54 2002
  DUMP: Average transfer rate: 517 KB/s
  DUMP: DUMP IS DONE
Kilobytes Out 959720
|------------------------------------------------
| Backup start: Sat Jun 01 02:04:22 2002
| Backup end:   Sat Jun 01 02:59:02 2002
|------------------------------------------------
At block 0.
|------------------------------------------------
| File number 2, index key 200206010203.38
| Backup of: /boot
| Date of this level 0 backup: Sat Jun 01 03:00:16 2002
| Date of last level 0 backup: the epoch
|------------------------------------------------
| (dump -0 -b 10 -a -f - /boot | gzip -4) | buffer -m 3m -s 10k -u 100 -t -p \
|  75 -B -o /dev/st0
|------------------------------------------------
  DUMP: Date of this level 0 dump: Sat Jun  1 03:00:16 2002
  DUMP: Date of last level 0 dump: the epoch
  DUMP: Dumping /dev/md0 (/boot) to standard output
  DUMP: Label: none
  DUMP: mapping (Pass I) [regular files]
  DUMP: mapping (Pass II) [directories]
  DUMP: estimated 2945 tape blocks.
  DUMP: Volume 1 started at: Sat Jun  1 03:00:17 2002
  DUMP: dumping (Pass III) [directories]
  DUMP: dumping (Pass IV) [regular files]
  DUMP: Volume 1 completed at: Sat Jun  1 03:00:20 2002
  DUMP: Volume 1 took 0:00:03
  DUMP: Volume 1 transfer rate: 979 KB/s
  DUMP: 2937 tape blocks (2.87MB)
  DUMP: finished in 3 seconds, throughput 979 KBytes/sec
  DUMP: Date of this level 0 dump: Sat Jun  1 03:00:16 2002
  DUMP: Date this dump completed:  Sat Jun  1 03:00:20 2002
  DUMP: Average transfer rate: 979 KB/s
  DUMP: DUMP IS DONE
Kilobytes Out 1740
|------------------------------------------------
| Backup start: Sat Jun 01 03:00:16 2002
| Backup end:   Sat Jun 01 03:00:49 2002
|------------------------------------------------
At block 0.
|------------------------------------------------
| Rewinding...
| Compressing log (all.0.20020601.gz)
| Linking all.latest.gz -> all.0.20020601.gz
|------------------------------------------------

File  Contents    (tape index 200206010203.38)
-----------------------------------------------
0  
1   level 0 / Sat Jun 01 02:04:22 2002 dump+gzip from e-smith-server
2   level 0 /boot Sat Jun 01 03:00:16 2002 dump+gzip from e-smith-server

Darrell May

Re: Backup
« Reply #1 on: June 03, 2002, 11:05:42 AM »
Something is very wrong as your tape position always shows, 'At block 0.'   So YES, if your backup is working at all, it appears /root is getting overwritten.  The tape position should of course change.  If you have never tested your backups I'd say you have either hardware problems or if the drive is confirmed good, hardware incompatibility with flexbackup.

Darrell

Tim Hogan

Re: Backup
« Reply #2 on: June 11, 2002, 08:04:43 AM »
Thanks for your response.  My backup drive is a Seagate STT22000A Hornet 10/20 GB drive.  I thought this model was compatible with a Mitel server.  I downloaded a Linux diagnostic program from Seagate and when I run it, it identifies the drive as follows:

ATAPE/IDE Logical Drive id:0
STT20000A, firmware 8A51 drive.  

I ran a rwtest with the diagnostic program and it claims to complete successfully.  I'm not quite sure where to go from here.  It is confusing to me that the diagnostics program identifies the drive as an STT20000A when the drive is labeled STT22000A.  I will try contacting Seagate for advice.

Tim Hogan

Re: Backup
« Reply #3 on: June 15, 2002, 08:08:01 AM »
I have contacted Seagate, and of course they say the blame is with the Software.  There response and my posting to them are shown below.  Is it possible I need to change some settings in the flexbackup configuration?

======================
Seagate Response
======================

For answers as to why a software program may overwrite previous
 sessions on a tape the manufacturer of the software must be contacted
 as the software issues the commands to the tape drive on which
 parts of the tape gets written to. This is not a function of the hardware
 as the hardware is waiting for commands from the software program.

 The tape diagnostic is seeing the tape drive correctly as all marketing
 numbers from the box are not imbedded in the firmware and the tape drive
 should identify as an STT20000A by the diagnostic and you are at the
 latest version of the software.

 David B.
 Seagate Technical Support

 ========== Customer's original message ==========


 ========================================================
 I have a STT22000A-RFT drive installed in a Linux server.  I have
 downloaded the Linux diagnostics and run them.  They seem to show no
 problems but identify the drive as an STT200000A, firmware 8A51.  I am
 using the tape provided with the drive.  When I run a backup each night it
 attempts to write everything at Block 0.  When I check the contents of the
 tape only the /boot directory appears (this is the last thing backed up).
 It appears everything gets backup up but since the drive writes everything
 to Block 0 everything before the last set is overwritten.  This is the
 message I receive:

For answers as to why a software program may overwrite previous
 sessions on a tape the manufacturer of the software must be contacted
 as the software issues the commands to the tape drive on which
 parts of the tape gets written to. This is not a function of the hardware
 as the hardware is waiting for commands from the software program.

 The tape diagnostic is seeing the tape drive correctly as all marketing
 numbers from the box are not imbedded in the firmware and the tape drive
 should identify as an STT20000A by the diagnostic and you are at the
 latest version of the software.

 David B.
 Seagate Technical Support

 ========== Customer's original message ==========


 ========================================================
 I have a STT22000A-RFT drive installed in a Linux server.  I have
 downloaded the Linux diagnostics and run them.  They seem to show no
 problems but identify the drive as an STT200000A, firmware 8A51.  I am
 using the tape provided with the drive.  When I run a backup each night it
 attempts to write everything at Block 0.  When I check the contents of the
 tape only the /boot directory appears (this is the last thing backed up).
 It appears everything gets backup up but since the drive writes everything
 to Block 0 everything before the last set is overwritten.  
=========================================================
See original posting for contents of the backup message I receive.

The contents of my flexbackup.conf file are:

#------------------------------------------------------------
# DO NOT MODIFY THIS FILE! It is updated automatically by the
# SME Server software. Instead, modify the source template in
# an /etc/e-smith/templates-custom directory. For more
# information, see http://www.e-smith.org/custom/
#
# copyright (C) 2002 Mitel Networks Corporation
#------------------------------------------------------------


$afio_echo_block = 'false';

$afio_compress_threshold = '3';

$afio_compress_cache_size = '2';


$blksize = '10';

$mt_var_blksize = 'true';


$buffer      = 'true';  
$buffer_megs = '3';    
$pad_blocks  = 'true';  


$compress    = 'gzip';  
$compr_level = '4';    


$cpio_format = 'newc';


$device    = '/dev/st0';

$dump_length = '0';

$dump_use_dumpdates = 'false';


$erase_rewind_only = 'false';


$erase_tape_all_level_zero = 'true';


$exclude_expr[0] = 'news/articles';
$exclude_expr[1] = '.*~$';  


$remoteshell = 'ssh';  
$verbose     = 'true';  
$sparse      = 'true';  
$indexes     = 'true';  


$logdir     = "/var/log/flexbackup";  
$stampdir = "/var/state/flexbackup";
$index    = "$stampdir/index";      
$comp_log = "gzip";                  
$prefix     = "";             
$sprefix  = "";                       
$tmpdir   = "/tmp";                  


$nocompress_types = 'mp3 Z z gz gif zip lha jpeg jpg taz tgz deb rpm bz2';


$tar_echo_record_num = 'false';

$tar_atime_preserve = 'true';


$traverse_fs = 'false';


$filesystems[0] = '/ /boot';


$type = 'dump';

Des Dougan

Re: Backup
« Reply #4 on: June 15, 2002, 10:34:14 AM »
Tim Hogan wrote:
>
> $device    = '/dev/st0';


Tim,

I think this is your problem - the device should be /dev/nst0 - the "n" signifies non-rewind. To change it:

/sbin/e-smith/db configuration setprop backup Device /dev/nst0
/sbin/e-smith/signal-event conf-backup

However, the block 0 part is also a problem. I have the same drive, and am waiting on feedback from the bugs list after I sent Mitel a trace of the dump activity. The dump is failing due the positioning at block 0, but if the backup is run again _without removing and re-inserting the tape_ the backup works successfully.

Des Dougan

Tim Hogan

Re: Backup
« Reply #5 on: June 16, 2002, 02:00:21 AM »
Thanks a million!  The backup now works and I was able to delete and restore a file successfully.

Tim Hogan

Des Dougan

Re: Backup
« Reply #6 on: June 16, 2002, 11:13:05 AM »
You're very welcome.

Des