> > Here is the output of df -k:
> > Filesystem 1024-blocks Used Available Capacity Mounted on
> > /dev/md0 5916736 811879 4798422 14% /
> > /dev/md1 3106031 41673 2903712 1% /var
> > /dev/sda1 83561123 11922926 67127549 15% /data
> >
> > To backup I use:
> > mt -f /dev/nst0 rewind
> > mf -f /dev/nst0 datcompression 2
> > dump 0uBf 96000000 /dev/nst0 /
> > dump 0uBf 96000000 /dev/nst0 /var
> > dump 0uBf 96000000 /dev/nst0 /data
> >
> > The third dump stops at 97% with "end of tape". The sum of the used blocks of
> > the filesystems is about 12 GB.
>
> That's right. It seems that you aren't really setting compression
> on, the math is clear :)
>
> I'd check the jumpers to see if the unit allows hardware
> compression to be set by software. Check either hp.com for this or the
> unit documentation. Once you're sure that the unit allows it, use mt-dds
> to set it. I found that mt-st don't do it properly...
(consulting www.hp.com): The dip switches are set to 'compression enabled at
power-on, with host control'. Without host control the result is the same. A
'dd if=/dev/zero of=/dev/tape' when dump asks for the next volume complains
about end of tape, too. 'mt -f /dev/nst0 tell' shows about 1,200,000 blocks
written. I hate it... How can I compress with SW when I use dump? With 'dump
0uBf 96000000 - | gzip | dd of=/dev/nst0' ?? IMHO restore -i will not work with
this...
Thomas