dump-users

Re: [Dump-users] -j/-z compression

[Stelian P. <st...@po...>]

On Thu, Oct 28, 2004 at 01:03:53PM +0200, Dick Visser wrote:

> On Thu, 28 Oct 2004, Stelian Pop wrote:
> 
> > Did you setup the blocksize to 0 using mt setblk / defblksize ?
> 
> FYI I use 37b dump on a debian system.
> 
> The default mt that come with is identifies like this:
> [root@aline backup]# mt --version
> GNU mt version 2.4.2
> 
> 
> However, there is also a package avaliabe called mt-st:
> [root@aline backup]# apt-get -s install mt-st
> Reading Package Lists... Done
> Building Dependency Tree... Done
> The following NEW packages will be installed:
>   mt-st
> 0 packages upgraded, 1 newly installed, 0 to remove and 51  not upgraded.
> Inst mt-st (0.7-2 Debian:3.0r3/stable)
> Conf mt-st (0.7-2 Debian:3.0r3/stable)
> 
> Might this be of any benefit?

YES.

GNU mt is a braindead tool, always use mt-st which is written by
the st driver maintainer.

Stelian.
-- 
Stelian Pop <st...@po...>    

Re: [Dump-users] -j/-z compression

[Dick V. <di...@ti...>]

On Thu, 28 Oct 2004, Stelian Pop wrote:

> Did you setup the blocksize to 0 using mt setblk / defblksize ?

FYI I use 37b dump on a debian system.

The default mt that come with is identifies like this:
[root@aline backup]# mt --version
GNU mt version 2.4.2


However, there is also a package avaliabe called mt-st:
[root@aline backup]# apt-get -s install mt-st
Reading Package Lists... Done
Building Dependency Tree... Done
The following NEW packages will be installed:
  mt-st
0 packages upgraded, 1 newly installed, 0 to remove and 51  not upgraded.
Inst mt-st (0.7-2 Debian:3.0r3/stable)
Conf mt-st (0.7-2 Debian:3.0r3/stable)

Might this be of any benefit?


-- 
  *    ***     Dick Visser         TIENHUIS Networking
 **   *   *    Marco Polostraat 234-3  P: +31206843731
  *   * ***    1056 DP Amsterdam       F: +31208641420
  *   * * *    The Netherlands         M: +31622698108
  *   **  *    WWW:             http://www.tienhuis.nl
  *   *   *    Email:         d.n...@ti...
  *   *   *    IP Phone:          sip:di...@ti...
 ***   ***     PGP-key: http://www.tienhuis.nl/gpg.txt
Live IPv6 webcam:  http://www.terena.nl/~dick/cam2.asx

Re: [Dump-users] -j/-z compression

[Dick V. <di...@ti...>]

On Thu, 28 Oct 2004, Stelian Pop wrote:

> Did you setup the blocksize to 0 using mt setblk / defblksize ?

Yes, both 0 and 1 give the same result...

-- 
  *    ***     Dick Visser         TIENHUIS Networking
 **   *   *    Marco Polostraat 234-3  P: +31206843731
  *   * ***    1056 DP Amsterdam       F: +31208641420
  *   * * *    The Netherlands         M: +31622698108
  *   **  *    WWW:             http://www.tienhuis.nl
  *   *   *    Email:         d.n...@ti...
  *   *   *    IP Phone:          sip:di...@ti...
 ***   ***     PGP-key: http://www.tienhuis.nl/gpg.txt
Live IPv6 webcam:  http://www.terena.nl/~dick/cam2.asx

Re: [Dump-users] -j/-z compression

[Stelian P. <st...@po...>]

On Thu, Oct 28, 2004 at 11:47:05AM +0200, Dick Visser wrote:

> So I guess I am out of luck?

Did you setup the blocksize to 0 using mt setblk / defblksize ?

Stelian.
-- 
Stelian Pop <st...@po...>    

[Dump-users] -j/-z compression

[Dick V. <di...@ti...>]

Hi guys

The dump manpage says my tapedrive has to support variable length block
writing if I want to use -j or -z compression. I cannot deduct this
from the tapedrive docs, so I just tried it out. This is what my box says:

  DUMP: Volume 1 started with block 1 at: Thu Oct 28 10:43:49 2004
  DUMP: write error 70 blocks into volume 1: Input/output error
  DUMP: Do you want to rewrite this volume?: ("yes" or "no")
	no
  DUMP: Do you want to start the next tape?: ("yes" or "no")
	no
  DUMP: The ENTIRE dump is aborted.

The goes for a DDS4 and a DLT-VS160 drive.
dmesg shows this with the DLT drive:

Oct 28 11:43:50 aline kernel: st0: Error with sense data: Current st0:
sense key Illegal Request
Oct 28 11:43:50 aline kernel: Additional sense: Invalid field in cdb

And with my DDS4 drive my kernel says:

st0: Write not multiple of tape block size.

So I guess I am out of luck?

-- 
  *    ***     Dick Visser         TIENHUIS Networking
 **   *   *    Marco Polostraat 234-3  P: +31206843731
  *   * ***    1056 DP Amsterdam       F: +31208641420
  *   * * *    The Netherlands         M: +31622698108
  *   **  *    WWW:             http://www.tienhuis.nl
  *   *   *    Email:         d.n...@ti...
  *   *   *    IP Phone:          sip:di...@ti...
 ***   ***     PGP-key: http://www.tienhuis.nl/gpg.txt
Live IPv6 webcam:  http://www.terena.nl/~dick/cam2.asx

[Dump-users] Marsha New Breed ARMM Equity Alert excrete

[Marsha W. <ar...@el...>]

Sun, 24 Oct 2004 19:02:53 -0700

Hot New Issue - merican Resource Management, Inc.
TICKER: ARMM

ARMM - Set to explode Monday Morning - Hot sto, ck
in a hot space

Opening-Price:0.06
Near-term_target:0.50
Long-term_target:0.60

Breaking News for ARMM - American Resource Management, Inc.

ARMM-launch new service (EffHVAC)

Efficiency Technologies, Inc.'s New Centrifugal Chiller
Efficiency and Management Tool Can Help Save Industry
Billions in Energy Costs

How many times have you seen good stocks but you couldn't
get your hands on them in the right moment? We are alerting
you to a special company with a unique product that is set
to be profiled by many newsletters in the next 5-10 days -
this is your chance to get in! 
  
Big Gains Expected Monday, Get It At Opening.

ALL TECHNICAL INDICATORS SAY - BUY ARMM @ up to 35 cents! 
  
Don't Delay Get ARMM. At Opening Monday Morning.


mortar I.nformation wi.thin this email con. tains forwa.
rdlook. ing sta. tements within the meaning of s. ect, ion 2,
7A of the Se, curities Ac, t of 19. 3. 3 and Sec.tion 21. B
of the S. ec. urit. ies E. xchange. Ac.t of 19. 34. A. ny st.
a tements t.hat express or involve discuss, ions with re,
spect to pred. ictions, assumptions or future events or
performance are not statements of historical fact. rambler

personage peltry villager odds

Re: [Dump-users] found an old posting

[Stelian P. <st...@po...>]

On Mon, Oct 25, 2004 at 01:43:44PM +0200, hpb wrote:
> hi,
> 
> I found an old posting. Is the adressed issue as dramatic as it sounds???
> 
> hpb
> 
> http://lwn.net/2001/0503/a/lt-dump.php3

See http://dump.sourceforge.net/isdumpdeprecated.html

Stelian.
-- 
Stelian Pop <st...@po...>    

Re: [Dump-users] Excluding files from restore -C

[Stelian P. <st...@po...>]

On Sun, Oct 17, 2004 at 09:45:03PM -0700, Kenneth Porter wrote:

> How can I tell restore to ignore files for a verify pass (-C)? I've blocked 
> them from dump using -E and a file containing a list of inodes. (I could 
> also use "chattr +d" to exclude them.) But the missing files are cluttering 
> up the verify so it's going to be hard to spot genuine verify errors.

Is the attached patch what you need ?


Index: dump/traverse.c
===================================================================
RCS file: /cvsroot/dump/dump/dump/traverse.c,v
retrieving revision 1.61
diff -u -r1.61 traverse.c
--- dump/traverse.c	1 Jul 2004 09:14:49 -0000	1.61
+++ dump/traverse.c	25 Oct 2004 10:38:28 -0000
@@ -985,6 +985,11 @@
 	struct direct *dp;
 	int reclen;
 
+	/* do not save entries to excluded inodes */
+	if (TSTINO(dirent->inode, dumpinomap) == 0 &&
+	    TSTINO(dirent->inode, dumpdirmap) == 0)
+		return 0;
+
 	p = (struct convert_dir_context *)private;
 
 	reclen = EXT2_DIR_REC_LEN((dirent->name_len & 0xFF) + 1);
-- 
Stelian Pop <st...@po...>    

[Dump-users] found an old posting

[hpb <hp...@ht...>]

hi,

I found an old posting. Is the adressed issue as dramatic as it sounds???

hpb

http://lwn.net/2001/0503/a/lt-dump.php3

I copy it below:-

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


From:	 Linus Torvalds <tor...@tr...>
To:	 Neil Conway <nco...@uk...>
Subject: Re: [PATCH] SMP race in ext2 - metadata corruption.
Date:	 Fri, 27 Apr 2001 09:59:46 -0700 (PDT)
Cc:	 Kernel Mailing List <lin...@vg...>


[ linux-kernel added back as a cc ]

On Fri, 27 Apr 2001, Neil Conway wrote:

 >>
 >> I'm surprised that dump is deprecated (by you at least ;-)).  What to
 >> use instead for backups on machines that can't umount disks regularly?


Note that dump simply won't work reliably at all even in 2.4.x: the buffer
cache and the page cache (where all the actual data is) are not
coherent. This is only going to get even worse in 2.5.x, when the
directories are moved into the page cache as well.

So anybody who depends on "dump" getting backups right is already playing
russian rulette with their backups.  It's not at all guaranteed to get the
right results - you may end up having stale data in the buffer cache that
ends up being "backed up".

Dump was a stupid program in the first place. Leave it behind.


 >> I've always thought "tar" was a bit undesirable (updates atimes or
 >> ctimes for example).


Right now, the cpio/tar/xxx solutions are definitely the best ones, and
will work on multiple filesystems (another limitation of "dump"). Whatever
problems they have, they are still better than the _guaranteed_(*)  data
corruptions of "dump".

However, it may be that in the long run it would be advantageous to have a
"filesystem maintenance interface" for doing things like backups and
defragmentation..

		Linus

(*) Dump may work fine for you a thousand times. But it _will_ fail under
the right circumstances. And there is nothing you can do about it.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to maj...@vg...
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: [Dump-users] GUI?

[Stelian P. <st...@po...>]

On Mon, Oct 25, 2004 at 11:08:45AM +0200, Dick Visser wrote:

> On Mon, 25 Oct 2004, Stelian Pop wrote:
> 
> > As you probably found out by the lack of responses, there is no GUI
> > available for restore.
> 
> Maybe it is not the right question here, but does anyone know of backup
> software that:
> 
> -runs on linux (duh)
> -can be restored with a GUI on MacOS/windows?

I sure there are a lot, but you can take a look at (the proprietary)
esbackup:
	http://www.ugsoft.de/intl/esbackup/index.html
which is based on dump (Uwe Gohlke, the esbackup author, has contributed
a lot of code back into the opensource dump/restore).

Stelian.
-- 
Stelian Pop <st...@po...>    

Re: [Dump-users] GUI?

[Dick V. <di...@ti...>]

On Mon, 25 Oct 2004, Stelian Pop wrote:

> As you probably found out by the lack of responses, there is no GUI
> available for restore.

Maybe it is not the right question here, but does anyone know of backup
software that:

-runs on linux (duh)
-can be restored with a GUI on MacOS/windows?

Preferrably opensource but if nothing decent is available, a commercial
solution would also suffice.

-- 
  *    ***     Dick Visser         TIENHUIS Networking
 **   *   *    Marco Polostraat 234-3  P: +31206843731
  *   * ***    1056 DP Amsterdam       F: +31208641420
  *   * * *    The Netherlands         M: +31622698108
  *   **  *    WWW:             http://www.tienhuis.nl
  *   *   *    Email:         d.n...@ti...
  *   *   *    IP Phone:          sip:di...@ti...
 ***   ***     PGP-key: http://www.tienhuis.nl/gpg.txt
Live IPv6 webcam:  http://www.terena.nl/~dick/cam2.asx

Re: [Dump-users] GUI?

[Stelian P. <st...@po...>]

On Thu, Oct 21, 2004 at 12:25:38PM +0200, Dick Visser wrote:

> Hi guys
> 
> As much as I love dump/restore as a sysadmin, some of my customers would
> like to be able to restore files themselves without using a terminal.
> This hold especially true for people that have linux servers running
> netatalk, where filenames often start with the ":" sign and have lots of
> spaces.
> 
> So: is there any GUI (either native compiled apps, PHP etc) available for
> dump/restore? I guess only the restore part would suffice.

As you probably found out by the lack of responses, there is no GUI
available for restore.

> If it is not available, I might consider building a PHP app myself.

It should not be too difficult... 

PHP is maybe not the most recommended language but hey, it's your
choice...

Stelian.
-- 
Stelian Pop <st...@po...>    

[Dump-users] GUI?

[Dick V. <di...@ti...>]

Hi guys

As much as I love dump/restore as a sysadmin, some of my customers would
like to be able to restore files themselves without using a terminal.
This hold especially true for people that have linux servers running
netatalk, where filenames often start with the ":" sign and have lots of
spaces.

So: is there any GUI (either native compiled apps, PHP etc) available for
dump/restore? I guess only the restore part would suffice.

If it is not available, I might consider building a PHP app myself.

Best regards,

-- 
  *    ***     Dick Visser         TIENHUIS Networking
 **   *   *    Marco Polostraat 234-3  P: +31206843731
  *   * ***    1056 DP Amsterdam       F: +31208641420
  *   * * *    The Netherlands         M: +31622698108
  *   **  *    WWW:             http://www.tienhuis.nl
  *   *   *    Email:         d.n...@ti...
  *   *   *    IP Phone:          sip:di...@ti...
 ***   ***     PGP-key: http://www.tienhuis.nl/gpg.txt
Live IPv6 webcam:  http://www.terena.nl/~dick/cam2.asx

[Dump-users] Excluding files from restore -C

[Kenneth P. <sh...@se...>]

How can I tell restore to ignore files for a verify pass (-C)? I've blocked 
them from dump using -E and a file containing a list of inodes. (I could 
also use "chattr +d" to exclude them.) But the missing files are cluttering 
up the verify so it's going to be hard to spot genuine verify errors.

Re: [Dump-users] restore incorrectly changes mode from 0777 to 0700 for the rip rescue distro

[Alan W. I. <ir...@be...>]

On 2004-09-23 07:20+0200 Stelian Pop wrote:

> On Wed, Sep 22, 2004 at 02:02:01PM -0700, Alan W. Irwin wrote:
>
>>> I see that rip exists in linux and BSD versions. You are using the Linux
>>> version, don't you ?
>>
>> Yes.  My latest tests were for RIP-11.0.iso.bin which is the Slackware-based
>> Linux version of RIP with kernel-2.6.8.
>>
>> Additional information is if I try to create symlinks by hand on rip they
>> come out as mode 755.  So it is not the same as the mode 700 created by
>> restore under RIP, but nevertheless it is different from the expected 777
>> (see above).
>
> Perhaps the umask is used ? If you set the umask to 0 do the symlinks
> get created with 0777 ?

It depends....  Here is the story in detail.

I tried changing the RIP umask from the default 0022 to 0000 and and it made
no difference to restore; restore continued to change all symlink
permissions to 0700. I also created a pristine ext3 filesystem (using
'mke2fs -j /dev/hda3' on RIP) which was mounted with 'mount /dev/hda3
/mnt/dum3'.  The results of the subsequent mount command to list out options
for each filesystem were as follows:

tmpfs on / type tmpfs (rw)
/dev/hda3 on /mnt/dum3 type ext3 (rw)

The first one is apparently how RIP creates a root file system in RAM, and
the second one shows no special options for /mnt/dum3

But if I am in the /mnt directory, i.e. in the tmpfs root filesystem, then the
series of commands:

umask 0000
ln -s /tmp/test_0000
umask 0022
ln -s /tmp/test_0022
umask 0077
ln -s /tmp/test_0077
ls -l test*

shows permissions of all links are 0777, but if I do the same series in
/mnt/dum3, i.e., in the ext3 filesystem then test_0000 has permissions of
0777, test_0022 has permissions of 0755, and test_0077 has permissions of
0700.  So the behaviour of symlink permissions depends on the filesystem for
RIP (and possibly Slackware since RIP is based on that).

I could find no option in mke2fs or tune2fs that would affect symlink
permissions.  It seems to me the only possibility left is there is a kernel
compile option or run-time parameter that is affecting the behaviour of
symlink permissions on ext2/3 filesystems.  The trouble is finding what the
option is and learning how to change it so that restore works correctly for
symlinks. Note this apparently only seems to be a problem for RIP and
perhaps other Slackware-based distros.  Debian-testing and knoppix (also
based on Debian) both seem fine.

My RIP, Debian testing, and knoppix testing has all been done for
kernel-2.6.  Does anybody on this list have access to a Slackware kernel-2.6
system to check whether restore is able to create the correct symlink
permissions?

Alan
__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the PLplot scientific plotting software
package (plplot.org), the Yorick front-end to PLplot (yplot.sf.net), the
Loads of Linux Links project (loll.sf.net), and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

Re: [Dump-users] restore incorrectly changes mode from 0777 to 0700 for the rip rescue distro

[Stelian P. <st...@po...>]

On Wed, Sep 22, 2004 at 02:02:01PM -0700, Alan W. Irwin wrote:

> >I see that rip exists in linux and BSD versions. You are using the Linux
> >version, don't you ?
> 
> Yes.  My latest tests were for RIP-11.0.iso.bin which is the Slackware-based
> Linux version of RIP with kernel-2.6.8.
> 
> Additional information is if I try to create symlinks by hand on rip they
> come out as mode 755.  So it is not the same as the mode 700 created by
> restore under RIP, but nevertheless it is different from the expected 777
> (see above).

Perhaps the umask is used ? If you set the umask to 0 do the symlinks
get created with 0777 ?

Stelian.
-- 
Stelian Pop <st...@po...>

Re: [Dump-users] restore incorrectly changes mode from 0777 to 0700 for the rip rescue distro

[Alan W. I. <ir...@be...>]

On 2004-09-22 21:57+0200 Stelian Pop wrote:

> On Mon, Sep 20, 2004 at 09:44:03PM -0700, Alan W. Irwin wrote:
>
>> This verification step shows many files have their modes
>> changed from 0777 to 0700.  Spot checks of a few of the thousands of
>> occurrences that appeared for a big filesystem showed symlinks were involved
> [...]
>
> I thought that on linux symlinks cannot have a mode different than 0777.

I thought that as well, but see below.

>
>> so I created a simple example filesystem with symlinks and demonstrated the
>> problem also with that simple example on rip.
>
> I see that rip exists in linux and BSD versions. You are using the Linux
> version, don't you ?

Yes.  My latest tests were for RIP-11.0.iso.bin which is the Slackware-based
Linux version of RIP with kernel-2.6.8.

Additional information is if I try to create symlinks by hand on rip they
come out as mode 755.  So it is not the same as the mode 700 created by
restore under RIP, but nevertheless it is different from the expected 777
(see above).

It turns out my restore scripts work fine for knoppix with no changes
to the symlink (or any other) modes.  So I have a workaround to the problem.
But knoppix is bloated and slow for rescue operations, and I prefer to use
RIP instead if possible so I have contacted Kent Robitti, the developer of
RIP, to see whether he knows of any kernel configuration or permissions
problem that could be causing the symlink mode inconsistency on RIP.

Because the list members here are presumbably quite experienced with restore
for rescue distros, I have posted my concerns here as well in case one of
you has run into this problem before (especially if you are using RIP or
slackware).  Also, somewhere in the recover code there must be a line that
creates symlinks and their permissions. What ext2/3 filesystem call does it
use?  There may be some configurable item associated with that call that can
be tracked down from this end.

Alan
__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the PLplot scientific plotting software
package (plplot.org), the Yorick front-end to PLplot (yplot.sf.net), the
Loads of Linux Links project (loll.sf.net), and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

Re: [Dump-users] Backup alteratives on Fedora

[Stelian P. <st...@po...>]

On Mon, Sep 20, 2004 at 05:23:11PM -0700, Kenneth Porter wrote:

> Interesting thread comparing dump to alternatives on the Fedora-devel list:
> 
> <http://thread.gmane.org/gmane.linux.redhat.fedora.devel/14220>

Interesting. Maybe we should work a bit on supporting xattr on
dump (we already have the format defined...).

Stelian.
-- 
Stelian Pop <st...@po...>

Re: [Dump-users] restore incorrectly changes mode from 0777 to 0700 for the rip rescue distro

[Stelian P. <st...@po...>]

On Mon, Sep 20, 2004 at 09:44:03PM -0700, Alan W. Irwin wrote:

> This verification step shows many files have their modes
> changed from 0777 to 0700.  Spot checks of a few of the thousands of
> occurrences that appeared for a big filesystem showed symlinks were involved
[...]

I thought that on linux symlinks cannot have a mode different than 0777.

> so I created a simple example filesystem with symlinks and demonstrated the
> problem also with that simple example on rip.

I see that rip exists in linux and BSD versions. You are using the Linux
version, don't you ?

Stelian.
-- 
Stelian Pop <st...@po...>

[Dump-users] restore incorrectly changes mode from 0777 to 0700 for the rip rescue distro

[Alan W. I. <ir...@be...>]

I need advice/help on dealing with a mode change problem I have encountered
with restored filesystems.  I am running the latest rip-11.0 rescue distro
and use the restore command to restore and verify filesystems that were
backed up with dump.  I have a script that does "restore -r ..." to restore
a filesystem followed by "restore -C ..." to verify it against the dump file
used to create it. This verification step shows many files have their modes
changed from 0777 to 0700.  Spot checks of a few of the thousands of
occurrences that appeared for a big filesystem showed symlinks were involved
so I created a simple example filesystem with symlinks and demonstrated the
problem also with that simple example on rip.  I then also did the same test
on Debian testing (with the same version 0.4b37 of dump/restore, same
script, and same dump file), and the problem disappears.

Therefore, I am pretty sure this is not a problem with restore, but instead
there exists some environmental (permission or whatever) problem which is
the cause of the mode change problems that I encounter on rip.  (Note, I
cannot abandon rescue distros like rip and use Debian testing instead,
because I need to restore my root file system on Debian testing.)

Has anyone on this list encountered a symlink mode problem like this before
that would force restore to change all symlink modes from 0777 to 0700?  For
example, I have heard of security exploits called symlink attacks, and I
wonder if rip does something special to stop those which is causing me grief
when I attempt to restore symlinks.  Another possibility to explain this
is some arcane permissions problem.

Of course, on linux, the mode is essentially ignored for all symlinks so
_if_ all occurrences of the problem are symlinks, then the problem is just
reduced to thousands of nuisance messages when I attempt to verify a large
filesystem using restore -C.  Nevertheless, I would like to get rid of those
messages for symlinks just in case there is a real file in the large
filesystem with mode 777 that has been incorrectly changed to 700 with real
consequences.

Alan
__________________________
Alan W. Irwin
email: ir...@be...
phone: 250-727-2902

Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).

Programming affiliations with the PLplot scientific plotting software
package (plplot.org), the Yorick front-end to PLplot (yplot.sf.net), the
Loads of Linux Links project (loll.sf.net), and the Linux Brochure Project
(lbproject.sf.net).
__________________________

Linux-powered Science
__________________________

[Dump-users] Backup alteratives on Fedora

[Kenneth P. <sh...@se...>]

Interesting thread comparing dump to alternatives on the Fedora-devel list:

<http://thread.gmane.org/gmane.linux.redhat.fedora.devel/14220>

Re: [Dump-users] Determine Dump Version from Tape

[Stelian P. <st...@po...>]

On Thu, Aug 05, 2004 at 02:51:23PM -0700, Peter Sangas wrote:

> Is there a way to determine what dump version was used to perform
> a dump to tape when all I have is the tape?

Unfortunately no. 

Why do you need that kind of information ? The latest version
of restore should be able to read all tape formats used by
previous versions of dump.

Stelian.
-- 
Stelian Pop <st...@po...>    

Re: [Dump-users] compress example is needed

[Stelian P. <st...@po...>]

On Tue, Jul 06, 2004 at 04:55:56PM +0400, gazolini wrote:

> Hi All!
> 
> Could anybody help me with an example of using compress parameter
> during dump operation? What are "compression" and "level" in -j and -z
> options?

There aren't two parameter but just one, called "compression level".
This is a one digit number, between 1 and 9, which has the same
meaning as the corresponding number in gzip or bzip2.

Stelian.
-- 
Stelian Pop <st...@po...>    

[Dump-users] compress example is needed

[gazolini <gaz...@in...>]

Hi All!

Could anybody help me with an example of using compress parameter
during dump operation? What are "compression" and "level" in -j and -z
options?

-- 
Best regards,
 gazolini                          mailto:gaz...@in...

Re: [Dump-users] Cannnot stat directory??

[Stelian P. <st...@po...>]

On Thu, Jul 01, 2004 at 07:39:16AM -0700, Ken Godee wrote:

> Ok, just a little confused in the restore "-D" option....
> 
> lets say  /dev/sda1 = /boot
> 
> mtab / fstab not available on rescue disk.
> 
> dump of /dev/sda1...
> 
> Dumping /dev/sda1 (unlisted file system) to /dev/nst0 on host adm380
> Label: /boot
> 
> Soooo....
> 
> Would I try  "-D /boot" ?

Yes, you have to mount /dev/sda1 as /boot (or /mnt/whatever), then
pass -D /mnt/whatever to restore.

> Still confused how it would match
> /boot to /dev/sda1 ?

Your confusion is coming from the fact that you ignore the fact
that restore does the comparision *against a mounted filesystem*

Although dump is able to dump a unmounted filesystem, restore is
not able to compare against the raw device.

> or is it just because I know that setting
> the counter "-s 1" is /dev/sda1
> ie. -D /boot .... -s 1

???? -s has nothing to do with that...

Stelian.
-- 
Stelian Pop <st...@po...>