FWIW department

 

We have a home directory with about 500GB of user data. I tried using the
multiple tapes setting but found that this was a hassle.

 

To get around the tape limit, I wrote a perl script that went thought the
/home directory and create files users-a users-b . users-z that contained
the names of users based on the first letter of their user name.  Each file
looks something like:

./qfang

./qiu

./qkhan

./qqu

./qwu

 

My disklist looks like the following:

 

zorn  /export/users-a  /export/fssnap/users  {

      users-tar

      include list "/local/Amanda/UsersList/users-a"

}

.

zorn  /export/users-a  /export/fssnap/users  {

      users-tar

      include list "/local/Amanda/UsersList/users-z"

}

 

# catch everything else. Not necessary now

#

zorn  /export/users-REST  /export/fssnap/users  {

     users-tar

     exclude append "./[a-z]*"

} 

 

I backup based on the letter of the persons first name. In my case I do not
have a group of users that have files greater than the size of one tape. It
happens that the file size of each group of people is different which makes
the backup go quicker, which is a side benefit.

 

Restore are also quick.

 

Robert

 

 

 

  _____  

From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
On Behalf Of FL
Sent: Wednesday, February 21, 2007 9:22 PM
To: [EMAIL PROTECTED]
Cc: [email protected]
Subject: Re: AMANDA reports no slots, but there ARE slots!

 

Now that this problem is overcome, have I run out of tape? Perhaps I need to
split directiories up (200 GB tape_

 

These dumps were to tape Daily-12.
*** A TAPE ERROR OCCURRED: [No more writable valid tape found].
Some dumps may have been left in the holding disk.
Run amflush to flush them to tape.
The next tape Amanda expects to use is: Daily-13. 

FAILURE AND STRANGE DUMP SUMMARY:
 wilf  /home/IOMEGANAS  lev 0  FAILED [out of tape]
 wilf  /home/IOMEGANAS  lev 0  FAILED [data write: Broken pipe]
 wilf  /home/IOMEGANAS  lev 0  FAILED [dump to tape failed] 


 

On 2/21/07, FL <[EMAIL PROTECTED]> wrote: 

 

On 2/21/07, Stefan G. Weichinger <[EMAIL PROTECTED] > wrote: 

FL schrieb:
> A host check reports no slots available, when there are slots available.
> What causes this, and what is the remedy? Thanks.
>
> Changer problem: no slots available

In my case it helped to change the group of the used changer-device: 

# chgrp tape /dev/sg2

when the amanda-user is member of the group tape. 

# ll /dev/sg2
crw-rw---- 1 root tape 21, 2 16. Feb 11:35 /dev/sg2

After that you can try

amanda ~ $ mtx -f /dev/sg2 status 

and amcheck should find slots then.

The group of the changer-device might be reset after reboot, due to udev 
and similar stuff, haven't yet researched how to fix that here. I do it
manually, one could also add this to some boot-script. 

Stefan

Hi,
In fact, that's what I've been doing, and I've added it to a boot script;
e.g.,

modprobe sg
chown root.tape /dev/sg2
if [ ! -h /dev/changer ]; then
   ln -s /dev/sg2 /dev/changer
fi

(My changer device is also /dev/sg2.) I must look into udev...I may be
attempting
to change ownership and set the symbolic link at the wrong time... 
FL

 

Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to