Amanda-Users

Re: short write even if the dumps are just 10% of the tape size

2005-01-17 07:56:47
Subject: Re: short write even if the dumps are just 10% of the tape size
From: Peter Guhl <pgnews AT siconline DOT ch>
To: Amanda Mailingliste <amanda-users AT amanda DOT org>
Date: Mon, 17 Jan 2005 13:45:04 +0100
Hi Paul

On Mon, 2005-01-17 at 12:38, Paul Bijnens wrote:
> But you can do just the same using dd:
> 
>     dd if=/dev/random of=/dev/st0 bs=32k
> 
> and count how many 32Kbyte blocks get on the tape.
> Expect about 15000000000 / 32768 or about 457763 blocks.
> 
> Remember, it takes about 2-3 hours to fill a tape.
> 
> Multiple runs should result in an almost constant number of blocks
> (plus/minus a few hundred).

root@:/usr/local/etc/amanda/backup> dd if=/dev/urandom of=/dev/nrsa0 bs=32k
dd: /dev/nrsa0: short write on tape device
1301+0 records in
1300+1 records out
42598476 bytes transferred in 13.372441 secs (3185542 bytes/sec)
root@:/usr/local/etc/amanda/backup> dd if=/dev/urandom of=/dev/nrsa0
bs=32k
dd: /dev/nrsa0: short write on tape device
1756+0 records in
1755+1 records out
57507912 bytes transferred in 18.072619 secs (3182046 bytes/sec)
root@:/usr/local/etc/amanda/backup> dd if=/dev/urandom of=/dev/nrsa0
bs=32k
dd: /dev/nrsa0: short write on tape device
127+0 records in
126+1 records out
4128840 bytes transferred in 1.362397 secs (3030570 bytes/sec)
root@:/usr/local/etc/amanda/backup> dd if=/dev/urandom of=/dev/nrsa0
bs=32k
dd: /dev/nrsa0: short write on tape device
1935+0 records in
1934+1 records out
63373386 bytes transferred in 20.628743 secs (3072092 bytes/sec)
root@:/usr/local/etc/amanda/backup> dd if=/dev/urandom of=/dev/nrsa0
bs=32k
dd: /dev/nrsa0: short write on tape device
354+0 records in
353+1 records out
11567158 bytes transferred in 3.702742 secs (3123944 bytes/sec)
root@:/usr/local/etc/amanda/backup> dd if=/dev/urandom of=/dev/nrsa0
bs=32k
dd: /dev/nrsa0: short write on tape device
704+0 records in
703+1 records out
23035980 bytes transferred in 7.960560 secs (2893764 bytes/sec)

It was a little bit faster ;-) but I guess that's mainly because this
drive isn't in a very good shape... at least those rapidly changing
numbers look a lot like that.

Regards
   Peter