Re: can't restore irix client with linux host

[Dietmar Goldbeck]

On Wed, Mar 06, 2002 at 09:06:04PM -0500, Jennifer Sturm wrote:
> Thanks, that works great. Is there a way to do an interactive restore 
> if I'm backing up with tar? I got it to restore by piping amrestore to 
> tar, but not interactively.
> 

GNUtar itself has no interactive restore, but you can use amrecover to
browse Amandas index and select files.  Amrecover works fine with
GNUtar.

-- 
 Alles Gute / best wishes  
 Dietmar GoldbeckE-Mail: [EMAIL PROTECTED]
Reporter (to Mahatma Gandhi): Mr Gandhi, what do you think of Western
Civilization?  Gandhi: I think it would be a good idea.

Re: tapetype and tapedev without tape drive

[Dietmar Goldbeck]

On Wed, Mar 06, 2002 at 03:22:06PM -0500, Jeffrey S. Auerbach wrote:
> Hello
> I am still waiting on the delivery of our tape drive.  For now I'd like
> to run AMANDA without a tape device.  According to the docs, this can be
> done but I am encountering some errors when I run amdump.
> 
> What values need to be entered for tapetype and/or tapedev?  Currently
> tapedev is /dev/null and tapetype is EXB-8500 in anticipation of our
> tape device.  I'd like to have my amanda.conf set for just keeping the
> backups on the holding disk for now. I believe that the rest of the file
> is set up correctly.
> 

Don't use /dev/null. 

This is for internal testing of Amanda only and really "backs up" to
/dev/null. tapedev and tapetype don't matter, otherwise. You can
safely use the real device (accessing /dev/nst0 under Linux behaves
like a real drive with no tape inserted, i don't know FreeBSD) or
something nonexistant like /not/found/tape

You have to set reserve in your amanda.conf to something below 100%

When your tape drive arrives, you can run amflush. 

  Ciao
Dietmar


-- 
 Alles Gute / best wishes  
 Dietmar GoldbeckE-Mail: [EMAIL PROTECTED]
Reporter (to Mahatma Gandhi): Mr Gandhi, what do you think of Western
Civilization?  Gandhi: I think it would be a good idea.

Re: can't restore irix client with linux host

[Jennifer Sturm]

Thanks, that works great. Is there a way to do an interactive restore 
if I'm backing up with tar? I got it to restore by piping amrestore to 
tar, but not interactively.

Thanks,

Jenn

- Original Message -
From: Dan Wilder <[EMAIL PROTECTED]>
Date: Wednesday, March 6, 2002 5:35 pm
Subject: Re: can't restore irix client with linux host

> On Wed, Mar 06, 2002 at 05:03:45PM -0500, Jenn Sturm wrote:
> > That should read "Now I've got a successful backup of the IRIX 
> box, 
> > but when I try to restore on the Linux host,..."
> 
> Dump and restore are architecture specific.  IRIX dumps must
> be restored on IRIX.  You may use Linux as the tape server.
> 
> For architecture-independent backup and restore use GNU tar.
> 
> > >I'm testing out backing  up an Irix machine with a Linux box. 
> I've 
> > >already successfully run and restored backups on the Linux 
box. 
> Now 
> > >I've got a successful backup of the linux box, but when I try to 
> > >restore using the following, I get this error:
> > >
> > >amrestore -p /dev/nst0 olympus /dev/dsk/dks3d11s3 | 
> /sbin/restore -ivf -
> > >
> > >Verify tape and initialize maps
> > >Input is from file/pipe
> > >amrestore: 0: skipping start of tape: date 20020305 label 
olympus00
> > >amrestore: 1: restoring 
olympus._dev_dsk_dks3d11s3.20020305.0
> > >/sbin/restore: Error reading dump file header
> > >
> > >This happens for every partition I try to restore.
> > >
> > >Any thoughts?
> > >
> > >Thanks,
> > >
> > >Jenn Sturm
> > >
> > >
> > >.---.
> > >|Jennifer Sturm |
> > >|System Administrator and Research Support Specialist   |
> > >|Chemistry Department   |
> > >|Hamilton College   |
> > >|   |
> > >|[EMAIL PROTECTED]|
> > >|315-859-4745   |
> > >|www.chem.hamilton.edu  |
> > >.---.
> > >|"Duct tape is like the force. It has a light side, |
> > >| and a dark side, and it holds the universe together." |
> > >|   -Carl Zwanzig   |
> > >|   |
> > >|   |
> > >|"One only needs two tools in life: WD-40 to make things|
> > >| go, and duct tape to make them stop. "|
> > >|   -G. Weilacher   |
> > >.---.
> > 
> 
> -- 
> -
> Dan Wilder <[EMAIL PROTECTED]>   Technical Manager & Editor
> SSC, Inc. P.O. Box 55549   Phone:  206-782-8808
> Seattle, WA  98155-0549URL 
http://embedded.linuxjournal.com/
> -
> 

Re: Inconsistent taping

[Jay Lessert]

On Wed, Mar 06, 2002 at 04:59:07PM -0500, Eric Trager wrote:
> I found out that I generally could get ~70 GB on a DLT-IV tape if I used
> hardware compression, so I set amanda's tape length to 60 GB and started
> backing up big partitions.
> 
> I was able to backup 55 and 52 GB partitions, but there's one 50 GB
> partition is particular that won't fit on a tape, apparently. I've
> attempted it several nights in a row, but with no success, even when
> amanda starts the tape with it.
> 
> Is what hardware compression can do with files variable? In this case,
> with the failure, it's a lot of oracle data, some of it is gzipped.

Sounds like you've answered your own question.  :-)

Your tape drive is either a DLT-7000 and has 35GB native capacity (or a
DLT-8000 with 40GB native).

If you use the HW compression device, and your data is compressible,
great, you get >35GB.  I've got my DLT-7000 set in my amanda.conf for
60GB.  Works great for me.

If your data is uncompressible, you're screwed.  Your capacity is
-guess what- 35GB (actually a little less).

You can try SW compression, in the hope that your data is somewhat
compressible and gzip can do sufficiently better (compared to the
tape drive HW) to squeeze it into 35GB.  You can dump/tar into
a gzip/wc pipe to check that out, something like:

% tar cf - /something | gzip -c --fast | dd of=/dev/null bs=1k
% tar cf - /something | gzip -c --best | dd of=/dev/null bs=1k

If you can't get <35GB out of this, you'll have to break the the
file system up into multiple disklist entries, or buy a bigger
tape drive.

-- 
Jay Lessert   [EMAIL PROTECTED]
Accelerant Networks Inc.   (voice)1.503.439.3461
Beaverton OR, USA(fax)1.503.466-9472

Re: can't restore irix client with linux host

[C. Chan]

Also Sprach Jenn Sturm:

> That should read "Now I've got a successful backup of the IRIX box,
> but when I try to restore on the Linux host,..."
>

The IRIX XFS filesystem uses xfsdump and xfsrestore. Linux XFS uses the same.
The /sbin/restore is for Linux ext2 dump/restore, which has had a rather
checkered history. I personally do not use Linux dump/restore for ext2.
xfsrestore for XFS seems to work the few times I have tested it, but
I'm staying with GNU tar on Linux for now.

> >--
> >
> >
> >I'm testing out backing  up an Irix machine with a Linux box. I've
> >already successfully run and restored backups on the Linux box. Now
> >I've got a successful backup of the linux box, but when I try to
> >restore using the following, I get this error:
> >
> >amrestore -p /dev/nst0 olympus /dev/dsk/dks3d11s3 | /sbin/restore -ivf -
> >
> >Verify tape and initialize maps
> >Input is from file/pipe
> >amrestore:   0: skipping start of tape: date 20020305 label olympus00
> >amrestore:   1: restoring olympus._dev_dsk_dks3d11s3.20020305.0
> >/sbin/restore: Error reading dump file header
> >
> >This happens for every partition I try to restore.
> >
> >Any thoughts?
> >
> >Thanks,
> >
> >Jenn Sturm
> >
> >
> >.---.
> >|Jennifer Sturm |
> >|System Administrator and Research Support Specialist   |
> >|Chemistry Department   |
> >|Hamilton College   |
> >|   |
> >|[EMAIL PROTECTED]|
> >|315-859-4745   |
> >|www.chem.hamilton.edu  |
> >.---.
> >|"Duct tape is like the force. It has a light side, |
> >| and a dark side, and it holds the universe together." |
> >|   -Carl Zwanzig   |
> >|   |
> >|   |
> >|"One only needs two tools in life: WD-40 to make things|
> >| go, and duct tape to make them stop. "|
> >|   -G. Weilacher   |
> >.---.
>
>
>

-- 
C. Chan < [EMAIL PROTECTED] >
PGP Public Key: finger [EMAIL PROTECTED]

Re: can't restore irix client with linux host

[Dan Wilder]

On Wed, Mar 06, 2002 at 05:03:45PM -0500, Jenn Sturm wrote:
> That should read "Now I've got a successful backup of the IRIX box, 
> but when I try to restore on the Linux host,..."

Dump and restore are architecture specific.  IRIX dumps must
be restored on IRIX.  You may use Linux as the tape server.

For architecture-independent backup and restore use GNU tar.

> >I'm testing out backing  up an Irix machine with a Linux box. I've 
> >already successfully run and restored backups on the Linux box. Now 
> >I've got a successful backup of the linux box, but when I try to 
> >restore using the following, I get this error:
> >
> >amrestore -p /dev/nst0 olympus /dev/dsk/dks3d11s3 | /sbin/restore -ivf -
> >
> >Verify tape and initialize maps
> >Input is from file/pipe
> >amrestore:   0: skipping start of tape: date 20020305 label olympus00
> >amrestore:   1: restoring olympus._dev_dsk_dks3d11s3.20020305.0
> >/sbin/restore: Error reading dump file header
> >
> >This happens for every partition I try to restore.
> >
> >Any thoughts?
> >
> >Thanks,
> >
> >Jenn Sturm
> >
> >
> >.---.
> >|Jennifer Sturm |
> >|System Administrator and Research Support Specialist   |
> >|Chemistry Department   |
> >|Hamilton College   |
> >|   |
> >|[EMAIL PROTECTED]|
> >|315-859-4745   |
> >|www.chem.hamilton.edu  |
> >.---.
> >|"Duct tape is like the force. It has a light side, |
> >| and a dark side, and it holds the universe together." |
> >|   -Carl Zwanzig   |
> >|   |
> >|   |
> >|"One only needs two tools in life: WD-40 to make things|
> >| go, and duct tape to make them stop. "|
> >|   -G. Weilacher   |
> >.---.
> 

-- 
-
 Dan Wilder <[EMAIL PROTECTED]>   Technical Manager & Editor
 SSC, Inc. P.O. Box 55549   Phone:  206-782-8808
 Seattle, WA  98155-0549URL http://embedded.linuxjournal.com/
-

Re: Inconsistent taping

[Brandon D. Valentine]

On Wed, 6 Mar 2002, Eric Trager wrote:

>Is what hardware compression can do with files variable? In this case,
>with the failure, it's a lot of oracle data, some of it is gzipped.

This is precisely why I do not use hardware compression.  The
effectiveness of the compression varies heavily based on what kind of
data you are backing up.  Chances are 50GB of gzip'd data is not going
to compress any at all.  I run a fast dual PIII as my tape server and it
handles doing the software compression for me just fine.  With software
compression planner's estimates are much more accurate and I sleep
easier at night knowing that my filesystems are reliably getting taped,
and if there is a problem that anything that wasn't taped was compressed
and stuck in my holding disk to be flushed the following morning.  My
holding disk probably wouldn't hold an entire run uncompressed, but with
software compression a $100 hard disk can hold an entire run.

-- 
Brandon D. Valentine <[EMAIL PROTECTED]>
Computer Geek, Center for Structural Biology

"This isn't rocket science -- but it _is_ computer science."
- Terry Lambert on [EMAIL PROTECTED]

Re: can't restore irix client with linux host

[Jenn Sturm]

That should read "Now I've got a successful backup of the IRIX box, 
but when I try to restore on the Linux host,..."

>--
>
>
>I'm testing out backing  up an Irix machine with a Linux box. I've 
>already successfully run and restored backups on the Linux box. Now 
>I've got a successful backup of the linux box, but when I try to 
>restore using the following, I get this error:
>
>amrestore -p /dev/nst0 olympus /dev/dsk/dks3d11s3 | /sbin/restore -ivf -
>
>Verify tape and initialize maps
>Input is from file/pipe
>amrestore: 0: skipping start of tape: date 20020305 label olympus00
>amrestore: 1: restoring olympus._dev_dsk_dks3d11s3.20020305.0
>/sbin/restore: Error reading dump file header
>
>This happens for every partition I try to restore.
>
>Any thoughts?
>
>Thanks,
>
>Jenn Sturm
>
>
>.---.
>|Jennifer Sturm |
>|System Administrator and Research Support Specialist   |
>|Chemistry Department   |
>|Hamilton College   |
>|   |
>|[EMAIL PROTECTED]|
>|315-859-4745   |
>|www.chem.hamilton.edu  |
>.---.
>|"Duct tape is like the force. It has a light side, |
>| and a dark side, and it holds the universe together." |
>|   -Carl Zwanzig   |
>|   |
>|   |
>|"One only needs two tools in life: WD-40 to make things|
>| go, and duct tape to make them stop. "|
>|   -G. Weilacher   |
>.---.


-- 





.---.
|Jennifer Sturm |
|System Administrator and Research Support Specialist   |
|Chemistry Department   |
|Hamilton College   |
|   |
|[EMAIL PROTECTED]|
|315-859-4745   |
|www.chem.hamilton.edu  |
.---.
|"Duct tape is like the force. It has a light side, |
| and a dark side, and it holds the universe together." |
|   -Carl Zwanzig   |
|   |
|   |
|"One only needs two tools in life: WD-40 to make things|
| go, and duct tape to make them stop. "|
|   -G. Weilacher   |
.---.

can't restore irix client with linux host

[Jenn Sturm]

-- 


I'm testing out backing  up an Irix machine with a Linux box. I've 
already successfully run and restored backups on the Linux box. Now 
I've got a successful backup of the linux box, but when I try to 
restore using the following, I get this error:

amrestore -p /dev/nst0 olympus /dev/dsk/dks3d11s3 | /sbin/restore -ivf -

Verify tape and initialize maps
Input is from file/pipe
amrestore:  0: skipping start of tape: date 20020305 label olympus00
amrestore:  1: restoring olympus._dev_dsk_dks3d11s3.20020305.0
/sbin/restore: Error reading dump file header

This happens for every partition I try to restore.

Any thoughts?

Thanks,

Jenn Sturm


.---.
|Jennifer Sturm |
|System Administrator and Research Support Specialist   |
|Chemistry Department   |
|Hamilton College   |
|   |
|[EMAIL PROTECTED]|
|315-859-4745   |
|www.chem.hamilton.edu  |
.---.
|"Duct tape is like the force. It has a light side, |
| and a dark side, and it holds the universe together." |
|   -Carl Zwanzig   |
|   |
|   |
|"One only needs two tools in life: WD-40 to make things|
| go, and duct tape to make them stop. "|
|   -G. Weilacher   |
.---.

Inconsistent taping

[Eric Trager]

I just wanted to find out if any of you have seen this.

I found out that I generally could get ~70 GB on a DLT-IV tape if I used
hardware compression, so I set amanda's tape length to 60 GB and started
backing up big partitions.

I was able to backup 55 and 52 GB partitions, but there's one 50 GB
partition is particular that won't fit on a tape, apparently. I've
attempted it several nights in a row, but with no success, even when
amanda starts the tape with it.

Is what hardware compression can do with files variable? In this case,
with the failure, it's a lot of oracle data, some of it is gzipped.

- -  -   --
Eric Trager

Re: [Amanda-users] amrecover: did not get a reserved port: 50059

[John Dalbec]

This is fixed in CVS.  The following patch fixed the problem for me, 
but YMMV.  In particular, if your firewall is picky about source ports 
for outgoing connections, it won't help you.  Note that this is *not*
the patch in CVS; OTOH it only affects one file rather than two.  It
also brings the behavior of --with-portrange in line with the comments
in the configure file.  (With this patch, the --with-portrange option
only affects server ports (that listen for connections), rather than 
also affecting client ports (that initiate connections).)
John

--- amanda-2.4.2p2/common-src/stream.c.orig Tue Feb 27 19:53:16 2001
+++ amanda-2.4.2p2/common-src/stream.c  Tue Dec 18 15:41:55 2001
@@ -214,10 +214,12 @@
  * to get the desired port, and to make sure we return a port that
  * is within the range it requires.
  */
+/*
 #ifdef PORTRANGE
 if (bind_portrange(client_socket, &claddr, PORTRANGE) == 0)
goto out;
 #endif
+*/
 
 if (bind_portrange(client_socket, &claddr, 512, IPPORT_RESERVED - 1) == 0)
goto out;

Jason Thomas wrote:
> 
> amrecover is broken it does this to me as well, and someone else also
> reported this problem.
> 
> just do your recover manually. use 'dd'
> 
> On Wed, Mar 06, 2002 at 01:56:58PM +0100, Jordi Vidal wrote:
> > Hi
> >   I cant use amrecover, it fails with a port problem:
> >
> > --
> > [root@tapeserver /root]# /usr/local/amanda/sbin/amrecover Test-CFG
> > AMRECOVER Version 2.4.2p2. Contacting server on tapeserver ...
> > amrecover: did not get a reserved port: 50059
> > --
> >
> > It is the same machine client and server. I compiled amanda with:
> >
> > ./configure   --prefix=/usr/local/amanda
> >   --with-user=amanda
> >   --with-group=disk
> >   --with-portrange=5,50100
> >   --with-udpportrange=512,520
> >
> > (BTW: Was udpportrange < 1024 a stupid setting?)

Re: file too large

[Frank Smith]

--On Wednesday, March 06, 2002 16:32:01 -0500 Charlie Chrisman 
<[EMAIL PROTECTED]> wrote:

> /-- countach.i /dev/sda3 lev 0 FAILED ["data write: File too large"]
>
> I get this for two of my clients?  what does this mean?

Probably that your Amanda server has a file size limit smaller than the
size of the filesystem you are trying to back up on the client.
   Try setting chunksize in your amanda.conf to something your server can
handle ( try 1 GB ).

Frank

--
Frank Smith[EMAIL PROTECTED]
Systems Administrator Voice: 512-374-4673
Hoover's Online Fax: 512-374-4501

Re: file too large

[Joshua Baker-LePain]

On Wed, 6 Mar 2002 at 4:32pm, Charlie Chrisman wrote

> /-- countach.i /dev/sda3 lev 0 FAILED ["data write: File too large"]
> 
> I get this for two of my clients?  what does this mean?
> 
FAQ.  Set your chunksize to something less then 2GB-32Kb.  1GB is fine -- 
there's no performance penalty.

-- 
Joshua Baker-LePain
Department of Biomedical Engineering
Duke University

RE: tapetype and tapedev without tape drive

[Joshua Baker-LePain]

On Wed, 6 Mar 2002 at 4:25pm, Jeffrey S. Auerbach wrote

> >the file: driver in your tapetype -- details in amanda(8).  
> 
> Not sure I understand, I can't find the correct driver, or do I create a
> tapetype called "disk" with the appropriate settings?
> 
Look in the "OUTPUT DRIVERS" section of the man page (amanda(8)) -- there 
are options for 'tapedev' that aren't actually a physical tape device.

But if it you're just waiting for your tape drive to arrive, I wouldn't 
bother with that.  Just define tapedev as whatever it will be and dump to 
disk (i.e. in degraded mode).

-- 
Joshua Baker-LePain
Department of Biomedical Engineering
Duke University

RE: tapetype and tapedev without tape drive

[Jeffrey S. Auerbach]

>the file: driver in your tapetype -- details in amanda(8).  

Not sure I understand, I can't find the correct driver, or do I create a
tapetype called "disk" with the appropriate settings?





-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]] On Behalf Of Joshua Baker-LePain
Sent: Wednesday, March 06, 2002 3:44 PM
To: Jeffrey S. Auerbach
Cc: [EMAIL PROTECTED]
Subject: Re: tapetype and tapedev without tape drive

On Wed, 6 Mar 2002 at 3:22pm, Jeffrey S. Auerbach wrote

> What values need to be entered for tapetype and/or tapedev?  Currently
> tapedev is /dev/null and tapetype is EXB-8500 in anticipation of our
> tape device.  I'd like to have my amanda.conf set for just keeping the
> backups on the holding disk for now. I believe that the rest of the
file
> is set up correctly.

If you want to use files on the holding disk as if they were tapes, use 
the file: driver in your tapetype -- details in amanda(8).  Or, just set

the tapedev to the correct device (/dev/nst0 or whatever) -- amanda will

spot that there's no drive there and dump to holding disk.  You'll need
to 
set reserve to a low number in order to get level 0s.

I believe having tapedev=/dev/null just throws the backups out.

> I am using version 2.4.3.b3 on FreeBSD 4.5 as I was told that I needed
> to use this version in order to backup with a tape drive.

That's "without" a tape drive I assume.  Yes, 2.4.3b includes the tapeio

functionality, the file: driver mentioned above.

-- 
Joshua Baker-LePain
Department of Biomedical Engineering
Duke University

file too large

[Charlie Chrisman]

/-- countach.i /dev/sda3 lev 0 FAILED ["data write: File too large"]

I get this for two of my clients?  what does this mean?

-- 
Charlie Chrisman
Business Development Director
(859) 514-7600
(859) 514-7601 Fax


³The Intelligent Way to Work²


Sent using the Entourage X Test Drive.

Re: [Amanda-users] [Fwd: Re: outgrown my one tape drive (DDS-4) solution; any recommendations?]

[Jason Thomas]

I was continually getting out of tape errors, so I told amanda the tape
was 2GB smaller than it really was. (1GB smaller didn't help still got
out of tape).

> When I am attempting to back up all that I would like to, I get out of
> tape errors.  My backup from last night just finished, and I put only
> the most critical unix f/s in disklist and only one of the 4 (the
> largest I think, I'll have to double check) windows shares.

Re: [Amanda-users] amrecover: did not get a reserved port: 50059

[Jason Thomas]

amrecover is broken it does this to me as well, and someone else also
reported this problem.

just do your recover manually. use 'dd'

On Wed, Mar 06, 2002 at 01:56:58PM +0100, Jordi Vidal wrote:
> Hi
>   I cant use amrecover, it fails with a port problem:
> 
> --
> [root@tapeserver /root]# /usr/local/amanda/sbin/amrecover Test-CFG
> AMRECOVER Version 2.4.2p2. Contacting server on tapeserver ...
> amrecover: did not get a reserved port: 50059
> --
> 
> It is the same machine client and server. I compiled amanda with:
> 
> ./configure   --prefix=/usr/local/amanda 
>   --with-user=amanda
>   --with-group=disk 
>   --with-portrange=5,50100 
>   --with-udpportrange=512,520
> 
> (BTW: Was udpportrange < 1024 a stupid setting?)

Re: tapetype and tapedev without tape drive

[Joshua Baker-LePain]

On Wed, 6 Mar 2002 at 3:22pm, Jeffrey S. Auerbach wrote

> What values need to be entered for tapetype and/or tapedev?  Currently
> tapedev is /dev/null and tapetype is EXB-8500 in anticipation of our
> tape device.  I'd like to have my amanda.conf set for just keeping the
> backups on the holding disk for now. I believe that the rest of the file
> is set up correctly.

If you want to use files on the holding disk as if they were tapes, use 
the file: driver in your tapetype -- details in amanda(8).  Or, just set 
the tapedev to the correct device (/dev/nst0 or whatever) -- amanda will 
spot that there's no drive there and dump to holding disk.  You'll need to 
set reserve to a low number in order to get level 0s.

I believe having tapedev=/dev/null just throws the backups out.

> I am using version 2.4.3.b3 on FreeBSD 4.5 as I was told that I needed
> to use this version in order to backup with a tape drive.

That's "without" a tape drive I assume.  Yes, 2.4.3b includes the tapeio 
functionality, the file: driver mentioned above.

-- 
Joshua Baker-LePain
Department of Biomedical Engineering
Duke University

tapetype and tapedev without tape drive

[Jeffrey S. Auerbach]

Hello
I am still waiting on the delivery of our tape drive.  For now I'd like
to run AMANDA without a tape device.  According to the docs, this can be
done but I am encountering some errors when I run amdump.

What values need to be entered for tapetype and/or tapedev?  Currently
tapedev is /dev/null and tapetype is EXB-8500 in anticipation of our
tape device.  I'd like to have my amanda.conf set for just keeping the
backups on the holding disk for now. I believe that the rest of the file
is set up correctly.

I am using version 2.4.3.b3 on FreeBSD 4.5 as I was told that I needed
to use this version in order to backup with a tape drive. 

Incidentally here is a snippet of the error that I am getting in my
logs:

driver: result time 30.226 from taper: TAPE-ERROR [rewinding tape:
Operation not supported by device]

Also, how can I change the DEFAULT_CONFIG variable?

Thanks for all the help.

Jeff Auerbach   

Failed and Strange Dump Details

[Davidson, Brian]

I consistantly get the following error backing up on a Sun client:

 db-test.em / lev 0 FAILED [mesg read: Connection reset by peer]

the other disks I backup on the same box work just fine:
DUMP SUMMARY: 
 DUMPER STATSTAPER STATS  
HOSTNAME DISKL ORIG-KB OUT-KB COMP% MMM:SS  KB/s MMM:SS  KB/s 
-- -  
 db-test.ems. /eec1   1  2 200.0   0:00   5.6   N/A   N/A   
 db-test.ems. -xport/home 1   166266200.0   0:07  90.9   N/A   N/A  

I am also excluding the above two disks from the / backup with and exclude
file:
 # cat /usr/local/lib/amanda/exclude.gtar 
 ./eec
 ./export/home

I've checked the FAQ but can't find anything related directly to the failure
message I'm getting above.  Any help would be greatly appreciated.

Thanks,   

Brian Davidson
11710 Plaza America Drive
Reston, Virginia 20190
703.261.4694
703.261.5086 Fax

[Fwd: Re: outgrown my one tape drive (DDS-4) solution; anyrecommendations?]

[Gene Matthews]

Sorry, forgot to cc the list with my reply do Johnannesgene

-Forwarded Message-

From: Gene Matthews <[EMAIL PROTECTED]>
To: Johannes Niess <[EMAIL PROTECTED]>
Subject: Re: outgrown my one tape drive (DDS-4) solution;  any recommendations?
Date: 06 Mar 2002 12:49:28 -0500

On Wed, 2002-03-06 at 10:39, Johannes Niess wrote:

> 
> Gene,
> 
> Don't spend money too fast.
> 
> So you're saying that the compressed size of your biggest disklist
> entry is over 20 GB on a level 0?

No, I don't think that is the case.

> 
> In this case you could make different disklist entries for all sub
> directories, creating new, (hidden $) shares (just for backup). This
> works best if the subdir's are somehow constant and no new sibblings
> are created. E. g. if you did backup \\nt\c$ , you could backup
> \\nt\c$\winnt , \\nt\c$\programms, etc.
> 
> 
> Or are you just getting "out of tape" errors?
> 

When I am attempting to back up all that I would like to, I get out of
tape errors.  My backup from last night just finished, and I put only
the most critical unix f/s in disklist and only one of the 4 (the
largest I think, I'll have to double check) windows shares.

It did complete this time.  It took almost 11 hours to run, though. 
I've got to figure out a way to get that down.  That was for a level 0
on the one windows partition.

I am seeing that the amount of data for these windows partitions
actually backed up (when it doesn't run out of tape) is larger than the
estimate received before it starts.  Sometimes 1.3 - 1.4 times, i think.


> Amanda will balance the backup levels for you over time. So after a
> few cycles the tape with a level 0 of \\nt (current configuration)
> will have (high numbered) incrementals of all other partitions. You
> could increase "dumpcycle".

Since this one worked, I'll try adding another one in tonights run and
see how that works.

Thanks,

Gene

> 
> HTH,
> 
> Johannes Nieß
> 
-- 
Gene Matthews
Matthews Midrange Consulting, Inc.
(678) 923-8327
(877) 882-6291 (toll free)
http://mmc-inc.com

Re: Fwd: help for tape changer problem

[Gene Heskett]

On Wednesday 06 March 2002 09:37 am, Zhen Liu wrote:

[snip]

>From: Zhen Liu <[EMAIL PROTECTED]>
>To: [EMAIL PROTECTED]
>Date: Mon, 04 Mar 2002 12:03:21 -0500
>
>Hi,
>
>I am working on the amanda backup project now, and I am a newbie for
>both amanda and linux... right now, I have some problem with our tape
>changer...SOS!!! & HELP!!!
>
>The OS I am working on is Linux RH7.1, the tape changer we have is
>Adic Fastor, Quantum 7000 tape drive.
>
>I guess that something must be wrong with my configuration...when i
>run amcheck i get "amcheck-server: could not get changer info: open:
>/dev/sg0: success"
>
>-relevant part from /etc/amanda/daily/amanda.conf
>
>runtapes 2
>tpchanger "chg-scsi"
>tapedev "0"
>changerfile "etc/amanda/DailySet1/changer.conf"
>changerdev "dev/sg0

I don't believe this is the correct device for the changer although it 
varies from mechanism to mechanism.  It also isn't required to define 
it twice as you have here, and in fact it could confuse things.

Some changers address the changer robot at the next device up in the 
scsi address scheme, and some will address it at Logical Unit 1 at the 
same bus ID number as the drive is, using lun 0 for the drive.

In order to see the rest of such a device that uses lun's other than 0, 
you will have to re-compile your kernel after turning on the 'SCAN ALL 
LUNS' option with a "make xconfig" or other method.  This will make the 
boot a little slower, particularly if the actual device address is 6 as 
it will scan all luns for all bus addresses, taking about 1 second per 
lun at 7 luns per address.  Because of this, I usually set the hardware 
address to whatever is the next one above any disk drives, purposely 
placing the disks first in the find sequence so they don't get 
re-addressed if the tape drive sneezes wrong.

Once that is done and the new kernel installed and the machine 
rebooted, you may then inspect (less /var/log/dmesg) the boot records 
to see what actual device has been assigned to the robot when it was 
found.  Insert that line in the changer.conf file, leaving the one in 
amanda.conf out completely.

>- /etc/amanda/DailySet1/changer.conf-
>number_configs   1
># eject  1
>sleep60
>cleanmax 1000
>changerdev   /dev/sg0
>
>#
># Next comes the data for drive 0
>#
>config   0
>drivenum 0
>dev  /dev/nst0
>startuse 0
>enduse   6
>statfile /etc/amanda/DailySet1/st0-slot
>cleancart-1
>cleanfile/etc/amanda/DailySet1/st0-clean
>usagecount   /etc/amanda/DailySet1/st0-totaltime
>
>
>So, I am totally lost...Help?

Do you keep a cleaning tape in the last numerical slot?  If so, reduce 
the enduse by one, and tell it the 'cleancart' is in that last slot.  
Amanda will then do it automaticly (or should anyway) when the drive 
says it needs cleaning.  And FWIW, cleaning tapes should not be reused 
1000 times as they will have become way too contaminated to be usefull 
after 5 or 6 passes at best.  But, amanda counts the number of times 
its been used for one cleaning cycle, and that normally is just a small 
fraction of the tape in the cartridge.  Some drives won't even rewind 
the cleaning tape, forcing a fresh tape when that one is used up.

-- 
Cheers, Gene

Re: Build dies trying to make amoverview

[Bernhard R. Erdmann]

> I am attempting to make amanda 2.4.2p2 on a NetBSD 1.5_ALPHA system
> The make dies with the following message:
> 
> cat amcheckdb.sh > amcheckdb
> chmod a+x amcheckdb
> cat amcleanup.sh > amcleanup
> chmod a+x amcleanup
> cat amdump.sh > amdump
> chmod a+x amdump
> make: don't know how to make amoverview. Stop
> *** Error code 1

Use GNU make (gmake)

Re: Tapetype: STD2401LW

[Bernhard R. Erdmann]

> Just retested last night, thought it might be useful for someone else.
> Kind of a bummer, a 20/40GB drive only able to write out 16.2GB... maybe
> I have something misconfigured?

You used hardware compression blowing up the random data tapetype
writes.

Re: Tapetype: STD2401LW

[Toomas Aas]

Hi [EMAIL PROTECTED]!

On  6 Mar 02 at 9:47 you wrote:

> Just retested last night, thought it might be useful for someone else. 
> Kind of a bummer, a 20/40GB drive only able to write out 16.2GB... maybe 
> I have something misconfigured? 

If the drive has hardware compression turned on, then tapetype can show 
the tape smaller than it actually is. The reason is that tapetype tests 
the tape by writing random data to it, and when the drive tries to 
compress the data it might actually expand.

If this is not in a FAQ somewhere, it IMHO should be.

--
Toomas Aas | [EMAIL PROTECTED] | http://www.raad.tartu.ee/~toomas/
* The secret of the universe is @*^^^&# NO CARRIER

Re: outgrown my one tape drive (DDS-4) solution; anyrecommendations?

[Jeffery Smith]

Hi,

I am using a Cybernetics CY-8960 stand alone tape drive.  It uses 8mm
Mammoth-2 cartridges which hold 60GB native / 150GB compressed.  Nightly
I backup 55 clients and it has performed reasonably well for me.  The
tapes are pretty expensive, but I think they are cheaper than DLT.

Jeff


On Wed, 2002-03-06 at 10:39, Johannes Niess wrote:
> Gene Matthews <[EMAIL PROTECTED]> writes:
> 
> > Hi all,
> > 
> > I hope this isn't considered off topic.  I have been running amanda now
> > for a month or so and all is going well (thanks to alot of help from
> > this list!).
> > 
> > My tape server is a Sun E250 (Solaris 8) with a single DDS-4 tape
> > drive.  I'm backing up approximately 15 clients (mostly solaris 8 /linux
> > - RH7.2) with three windows boxes.  The last windows box I added is
> > taking me over the limit.  It has 4 large partitions, the largest of
> > which can't ever seem to fit on the tape with everything else when it
> > does a level 0 dump.
> > 
> > I don't really want to go the manual changer route since the backups are
> > cron'd and nobody is around in the middle of the night to change tapes.
> > 
> > I'm starting to look at another tape drive solution and leaning towards
> > DLT.  I've seen references to AIT-2/AIT-3 also; I'm not sure what that
> > is.  I'm not up on my tape technology I guess.
> > 
> > I'm thinking of a DLT stacker or maybe a low end DLT tape library.  Does
> > anyone have any recommendations that they know works well with Amanda? 
> > There seems to be plenty of h/w to choose from based upon my web
> > searches and looking at vendor sites, but I don't want to purchase
> > something that isn't going to work with Amanda.
> > 
> > I couldn't find any info like this in the FAQ and my search of the
> > archives didn't turn up anything either.
> > 
> > All input very much appreciated!
> > 
> > Thanks,
> > 
> > Gene
> 
> Gene,
> 
> Don't spend money too fast.
> 
> So you're saying that the compressed size of your biggest disklist
> entry is over 20 GB on a level 0?
> 
> In this case you could make different disklist entries for all sub
> directories, creating new, (hidden $) shares (just for backup). This
> works best if the subdir's are somehow constant and no new sibblings
> are created. E. g. if you did backup \\nt\c$ , you could backup
> \\nt\c$\winnt , \\nt\c$\programms, etc.
> 
> 
> Or are you just getting "out of tape" errors?
> 
> Amanda will balance the backup levels for you over time. So after a
> few cycles the tape with a level 0 of \\nt (current configuration)
> will have (high numbered) incrementals of all other partitions. You
> could increase "dumpcycle".
> 
> HTH,
> 
> Johannes Nieß

Tape type

[Arunav Mandal]

Hi,

Can anybody know what are the tape type parameters for Exabyte 210 tape
libary having Exb-8900 tape drive?

Arunav Mandal.

Re: outgrown my one tape drive (DDS-4) solution; any recommendations?

[Johannes Niess]

Gene Matthews <[EMAIL PROTECTED]> writes:

> Hi all,
> 
> I hope this isn't considered off topic.  I have been running amanda now
> for a month or so and all is going well (thanks to alot of help from
> this list!).
> 
> My tape server is a Sun E250 (Solaris 8) with a single DDS-4 tape
> drive.  I'm backing up approximately 15 clients (mostly solaris 8 /linux
> - RH7.2) with three windows boxes.  The last windows box I added is
> taking me over the limit.  It has 4 large partitions, the largest of
> which can't ever seem to fit on the tape with everything else when it
> does a level 0 dump.
> 
> I don't really want to go the manual changer route since the backups are
> cron'd and nobody is around in the middle of the night to change tapes.
> 
> I'm starting to look at another tape drive solution and leaning towards
> DLT.  I've seen references to AIT-2/AIT-3 also; I'm not sure what that
> is.  I'm not up on my tape technology I guess.
> 
> I'm thinking of a DLT stacker or maybe a low end DLT tape library.  Does
> anyone have any recommendations that they know works well with Amanda? 
> There seems to be plenty of h/w to choose from based upon my web
> searches and looking at vendor sites, but I don't want to purchase
> something that isn't going to work with Amanda.
> 
> I couldn't find any info like this in the FAQ and my search of the
> archives didn't turn up anything either.
> 
> All input very much appreciated!
> 
> Thanks,
> 
> Gene

Gene,

Don't spend money too fast.

So you're saying that the compressed size of your biggest disklist
entry is over 20 GB on a level 0?

In this case you could make different disklist entries for all sub
directories, creating new, (hidden $) shares (just for backup). This
works best if the subdir's are somehow constant and no new sibblings
are created. E. g. if you did backup \\nt\c$ , you could backup
\\nt\c$\winnt , \\nt\c$\programms, etc.


Or are you just getting "out of tape" errors?

Amanda will balance the backup levels for you over time. So after a
few cycles the tape with a level 0 of \\nt (current configuration)
will have (high numbered) incrementals of all other partitions. You
could increase "dumpcycle".

HTH,

Johannes Nieß

Re: Backups running fine, but getting empty reports

[Graham Dunn]

On Wed, Mar 06, 2002 at 08:48:48AM +0100, Johannes Niess wrote:
> Graham Dunn <[EMAIL PROTECTED]> writes:
> 
> > The body of the daily backup report is always empty. Is there a
> > /tmp/amanda/ file I can poke into to see if there's some file
> > permission problem?
> > 
> > Graham
> 
> Graham,
> 
> Did you change your "columnspec" parameter? Especially "-1" values
> seem to be buggy in 2.4.2.
> 
> I'd do an amreport and check if it segfaults. Database entries and
> recovery are still ok.

This is very strange now. I noticed last night that if I changed the
logdir parameter to

logdir   "/usr/adm/amanda/test/"

removing the "log" from the end, I could manually get a proper report
from amreport. Oh-ho, I thought. Problem fixed. However, this morning,
I go to check, and my /usr/adm/amanda/log directory is now called
/usr/adm/amanda/log.error.0/

So obviously I don't know as much as I think I do :]

(my columnspec hasn't been modified from the default

# columnspec "Disk=1:18,HostName=0:10,OutKB=1:7"

The log files from last night's run are 0 length, but using one from a
previous run results in:

To: operator
Subject: test AMANDA MAIL REPORT FOR March 2, 2002

These dumps were to tape test3.
The next tape Amanda expects to use is: test5.


STATISTICS:
  Total   Full  Daily
      
Estimate Time (hrs:min)0:00
Run Time (hrs:min) 0:07
Dump Time (hrs:min)0:03   0:03   0:00
Output Size (meg) 292.5  292.50.0
Original Size (meg)   292.5  292.50.0
Avg Compressed Size (%) -- -- --
Filesystems Dumped8  8  0
Avg Dump Rate (k/s)  1504.2 1504.2--

Tape Time (hrs:min)0:06   0:06   0:00
Tape Size (meg)   292.8  292.80.0
Tape Used (%)   9.09.00.0
Filesystems Taped 8  8  0
Avg Tp Write Rate (k/s)   779.1  779.1--

^L
NOTES:
  planner: Adding new disk shizuka:da0s1a.
  planner: Adding new disk shizuka:/home.
  planner: Adding new disk ns1.inscriber.com:da0s1a.
  planner: Adding new disk ns1.inscriber.com:/var/named.
  taper: tape test3 kb 299872 fm 8 [OK]

^L
DUMP SUMMARY:
 DUMPER STATSTAPER
STATS
HOSTNAME DISKL ORIG-KB OUT-KB COMP% MMM:SS  KB/s MMM:SS
KB/s
-- -

axp500   da0a0   45070  45120   --0:192404.0   0:56
809.6
ns1.inscribe /var/named  0  60 96   --0:00 180.5   0:01
68.6
ns1.inscribe da0s1a  0   39640  39680   --0:192056.0   0:54
729.0
oshaberi -al/mailman 08040   8096   --0:042013.4   0:11
742.5
oshaberi ad0s1a  0   36200  36256   --0:142655.9   0:48
748.9
shizuka  /home   0   98490  98528   --1:271130.7   2:01
812.4
shizuka  da0s1a  0   35100  35136   --0:36 973.4   0:44
806.1
tsuyoi   ad0s1a  0   36920  36960   --0:201848.6   0:49
752.0

(brought to you by Amanda version 2.4.3b2)

Tapetype: STD2401LW

[mark . hazen]

Just retested last night, thought it might be useful for someone else. 
Kind of a bummer, a 20/40GB drive only able to write out 16.2GB... maybe 
I have something misconfigured? 

Tapetype for a Seagate STD2401LW DSS4 DAT Drive, using 150m tapes (also 
marketed by IBM as IBM's 20/40GB, DDS4, 4MM Internal Tape Drive): 

define tapetype STD2401LW { 
comment "Seagate STD2401LW DSS4 20/40GB DAT Drive" 
length 16564 mbytes 
filemark 0 kbytes  
speed 2288 kps 
} 

-- 
Mark Hazen   DataBuilt, Inc.  (843) 836-2101 Ext. 251 

They may forget what you said, but they will never forget 
how you made them feel. --Carl W. Buechner 

Compaq TL891 on Solaris

[Roland Barmettler]

Hi

We're currently using Amanda 2.4.1p1 on Solaris 2.5.1 with a Compaq
TL891 DLT library. But we were unable to get the changer working with
mtx (current version) and sst, so we have to change tapes manually.
(that's why we have a library...  *ç%&$!)

Is anyone using successfully the same Library on Solaris with MTX and
Amanda ? I'm greatful for any suggestions. :-)

Greetings, Roland

--- bbp - Biveroni Batschelet Partners AG 
 Bahnhofstrasse 28, CH-5401 Baden, Switzerland
   Roland Barmettler <[EMAIL PROTECTED]>  --  Phone +41-56-2039602
--

outgrown my one tape drive (DDS-4) solution; any recommendations?

[Gene Matthews]

Hi all,

I hope this isn't considered off topic.  I have been running amanda now
for a month or so and all is going well (thanks to alot of help from
this list!).

My tape server is a Sun E250 (Solaris 8) with a single DDS-4 tape
drive.  I'm backing up approximately 15 clients (mostly solaris 8 /linux
- RH7.2) with three windows boxes.  The last windows box I added is
taking me over the limit.  It has 4 large partitions, the largest of
which can't ever seem to fit on the tape with everything else when it
does a level 0 dump.

I don't really want to go the manual changer route since the backups are
cron'd and nobody is around in the middle of the night to change tapes.

I'm starting to look at another tape drive solution and leaning towards
DLT.  I've seen references to AIT-2/AIT-3 also; I'm not sure what that
is.  I'm not up on my tape technology I guess.

I'm thinking of a DLT stacker or maybe a low end DLT tape library.  Does
anyone have any recommendations that they know works well with Amanda? 
There seems to be plenty of h/w to choose from based upon my web
searches and looking at vendor sites, but I don't want to purchase
something that isn't going to work with Amanda.

I couldn't find any info like this in the FAQ and my search of the
archives didn't turn up anything either.

All input very much appreciated!

Thanks,

Gene

-- 
Gene Matthews
Matthews Midrange Consulting, Inc.
(678) 923-8327
(877) 882-6291 (toll free)
http://mmc-inc.com

Fwd: help for tape changer problem

[Zhen Liu]

Zhen Liu
Via Webmail
[EMAIL PROTECTED]
-- Forwarded message ---


From: Zhen Liu <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Date: Tue, 05 Mar 2002 11:26:40 -0500

Zhen Liu
Via Webmail
[EMAIL PROTECTED]
-- Forwarded message ---


From: Zhen Liu <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Date: Mon, 04 Mar 2002 12:03:21 -0500

Hi,

I am working on the amanda backup project now, and I am a newbie for
both amanda and linux... right now, I have some problem with our tape
changer...SOS!!! & HELP!!!

The OS I am working on is Linux RH7.1, the tape changer we have is
Adic Fastor, Quantum 7000 tape drive.

I guess that something must be wrong with my configuration...when i
run amcheck i get "amcheck-server: could not get changer info: open:
/dev/sg0: success"

-relevant part from /etc/amanda/daily/amanda.conf

runtapes 2
tpchanger "chg-scsi"
tapedev "0"
changerfile "etc/amanda/DailySet1/changer.conf"
changerdev "dev/sg0

- /etc/amanda/DailySet1/changer.conf-
number_configs   1
# eject  1
sleep60
cleanmax 1000
changerdev   /dev/sg0

#
# Next comes the data for drive 0
#
config   0
drivenum 0
dev  /dev/nst0
startuse 0
enduse   6
statfile /etc/amanda/DailySet1/st0-slot
cleancart-1
cleanfile/etc/amanda/DailySet1/st0-clean
usagecount   /etc/amanda/DailySet1/st0-totaltime


So, I am totally lost...Help?

Thanks...


Zhen Liu
Via Webmail
[EMAIL PROTECTED]

RE: amanda cron job not working

[Thomas Robinson]

Sorry about being so generic: fails means I don't get any output in my
mailbox (where I usually get the amanda reports).

I've done a bit more sifting through the logs and come up with the
following in /var/log/messages:

Feb 28 09:41:52 geko60 xinetd[656]: Service amanda missing attribute
user
Feb 28 09:41:52 geko60 xinetd[656]: Service amandaidx missing attribute
user
Feb 28 09:41:52 geko60 xinetd[656]: Service amidxtape missing attribute
user

This was a fault where our NIS client wasn't started before xinetd.

I've restarted both ypbind and xinetd and rescheduled amcheck for a
test. I got some mail with the amcheck output as I would normally expect
so I think this solves the problem.

BTW regarding my other post - I'm using the following in
/etc/samba/smb.conf now:

character set = ISO8859-15
client code page = 850

and that seems to help smbclient translate most filenames correctly now.
(I only have one file now that doesn't translate properly - it has
comma's in the filename). I did notice that the W2K clients are all set
to codepage 437 but I can't find anyway to change it permanently.

Thanks for the tips.

Tom


> -Original Message-
> From: Paul Bijnens [mailto:[EMAIL PROTECTED]]
> Sent: 06 March 2002 11:12
> To: Thomas Robinson
> Subject: Re: amanda cron job not working
> 
> 
> 
> 
> Thomas Robinson wrote:
> 
> > The backup was recovered manually with amflush and I have since run
> > amcheck and amdump manually without problems. The only 
> thing is that the
> > original cron scheduling fails. I have restarted cron and checked a
> 
> I would help if you could elaborate on "fails".
> What makes you believe it fails?
> What error messages do you get?
> Are it cron errors or Amanda errors?
> etc
> 
> -- 
> Paul Bijnens, Lant Tel  +32 
> 16 40.51.40
> Interleuvenlaan 15 H, B-3001 Leuven, BELGIUM   Fax  +32 
> 16 40.49.61
> http://www.lant.com/   email:  
> [EMAIL PROTECTED]
> **
> *
> * I think I've got the hang of it now:  exit, ^D, ^C, ^\, ^Z, 
> ^Q, F6, *
> * quit,  ZZ, :q, :q!,  M-Z, ^X^C,  logoff, logout, close, 
> bye,  /bye, *
> * stop, end, F3, ~., ^]c, +++ ATH, disconnect, halt,  abort,  
> hangup, *
> * PF4, F20, ^X^X, :D::D, KJOB, F14-f-e, F8-e,  kill -1 $$,  
> shutdown, *
> * kill -9 1,  Alt-F4,  Ctrl-Alt-Del,  AltGr-NumLock,  Stop-A, 
>  ...*
> * ...  "Are you sure?"  ...   YES   ...   Phew ...   I'm out  
> *
> **
> *
> 

amanda cron job not working

[Thomas Robinson]

Hello,

I'm administering an amanda setup that uses cron to schedule the
backups. Everything was working fine until someone else forgot to change
the tape and amanda kicked up a fuss because the label was incorrect. (I
was out of the office for a week). The backup was sent to a holding
disk, but that was too small for all the backup data.

The backup was recovered manually with amflush and I have since run
amcheck and amdump manually without problems. The only thing is that the
original cron scheduling fails. I have restarted cron and checked a
number of other services for conflicts. I have other cron jobs that do
execute when scheduled but as yet amanda won't play with the other kids
in the cron sandbox...

What am I missing? Has anyone seen this behaviour before. Any
suggestions are appreciated

Thanks,

Tom

Thomas Robinson
Ehbas Ltd
T: 01273 234 665
F: 01273 704 499