Amanda-Users

Re: amtapetype doesn´t work

2005-11-10 09:24:38
Subject: Re: amtapetype doesn´t work
From: Jon LaBadie <jon AT jgcomp DOT com>
To: amanda-users AT amanda DOT org
Date: Thu, 10 Nov 2005 09:11:22 -0500
On Thu, Nov 10, 2005 at 12:45:19PM -0000, knappenschaenke wrote:
> --- In amanda-users AT yahoogroups DOT com, Jon LaBadie <jon@j...> wrote:
> >
> > So a write comparable to your tar command succeeded.  Then amanda
> > rewound the tape and started a second write.  That is when it 
> failed.
> > Perhaps you could run a test that more closely matches amanda's 
> code
> > by putting the following into a shell script and running it.
> > 
> >   mt -f /dev/nst0 rewind
> >   dd bs=32k if=/dev/random of=/dev/nst0 count=4000
> >   mt -f /dev/nst0 rewind
> >   dd bs=32k if=/dev/random of=/dev/nst0 count=4000
> > 
> > It looks to me like the installation, or the configuration,
> > of the drive is the problem.
> >
> 
> --------------------------------------------
> First
> --------------------------------------------
> asterix: # mt -f /dev/nst0l rewind
> asterix: # dd bs=32k if=/dev/random of=/dev/nst0 count=1000
> 0+1000 Datensätze ein
> 0+1000 Datensätze aus
> 13917 bytes (14 kB) copied, 5149,72 seconds, 0,0 kB/s
> 
> ---------------------------------------------
> Second
> ---------------------------------------------
> asterix: # mt -f /dev/nst0l rewind
> asterix: # dd bs=32k if=/dev/random of=/dev/nst0 count=1000
> 0+1000 Datensätze ein
> 0+1000 Datensätze aus
> 12938 bytes (14 kB) copied, 5764,14 seconds, 0,0 kB/s
> 
> So I didn´t test it with count=4000 !

I failed to explain why I wanted it in a shell script.
I specifically did not want the delay that is caused
by a human typing in the commands.  I sometimes see
with my drive (a dds3) a return from a rewind before
the drive is really ready for the next command.  There
was a chance that your drive and driver behaves similarly.
I need a delay after a rewind.

HOWEVER, your dd commands show something is seriously wrong
having nothing to do with amanda.

The bs=32k option set the block size to 32 KiloBytes each,
and count option asked dd to write 1000 of them for a total
of 32 MegaBytes.  Yet the report from dd shows that not even 
a single complet block (0+...) blocks were written.  Only
14 KBytes total in what, over 5000 seconds?  Somethings wrong.


-- 
Jon H. LaBadie                  jon AT jgcomp DOT com
 JG Computing
 4455 Province Line Road        (609) 252-0159
 Princeton, NJ  08540-4322      (609) 683-7220 (fax)

<Prev in Thread] Current Thread [Next in Thread>