Dear All,
Please tell me when it becomes OT or boring.
Gene Heskett wrote:
> Not this particular version. But it comes to mind that maybe
> there is a
> switch or jumper configuration option on the drive that
> either locks it
> on, or locks the config page in the flash from being changed from the
> factory approved defaults. Some sort of insurance against
> Joe SixPack
> turning their inventory into a brick to be warrantied.
Unfortunately there is no switch/jumper - I would really spend the effort to
dismount and reintegrate the drive to solve my problem. And it wasn't me who
bought the drive, I just have to deal with it.
Matthias Andree gave me some good advice here to use some linux scsi tools
to circumvent the error message of "mt", which could not write the mode page
(datcompression off).
As I wrote, I was able to change the compression using the sginfo command
(see my posting some days ago)... but ...
But afterwards I started the amtapetype to extract the drive capacity
information and please have a look what it reported:
Estimate phase 1...
Writing 8 Mbyte compresseable data: 708 sec
Estimate phase 2...
Writing 8 Mbyte uncompresseable data: 710 sec
Estimated time to write 2 * 20480 Mbyte: 3635200 sec = 1009 h 46 min
Man, 1009 hours :-( I think that by changing the mode page I slowed down the
drive dramatically.
[Quote] If amtapetype really runs for 9 days, you can be pretty sure there
is something wrong with your approach.[/quote] from:
http://www.amanda.org/docs/topten.html#id2552780
I still have a backup on the holding disk and some hours ago I started the
flush to the tape, which should take 3-4 hours for 20g usually. I will see
if it is still running tomorrow. If I do not find a solution I will have to
restore the manufacturer settings with 'compr. on' again.
> Because of this, one must read (dd to the rescue) that label
> block out
> to a scratch file, rewind the tape, turn the compression and the
> datcompression off, and before the tape is moved again, write that
> scratch file back to the tape with dd again, using the non-rewinding
> tape device so the tape isn't rewound and re-read, followed
> by several
> megabytes of /dev/zero (dd if=/dev/zero) to force the drive to flush
> its buffers to the media, at which point that compression flag in the
> header will finally be turned off. Whatever is on the tape will be
> toast of course, so do this immediately before the tape is to be
> re-used. Once you cycled through all the tapes, things
> should be fine
> till the drive (or a tape) is ~30~, fini, worn out, kaput.
Ya Gene, I am aware about this. I would happily spent this effort to have a
good drive until it will be kaputt :-)
Thank you very much and best regards
Uwe
____________
Virus checked by G DATA AntiVirusKit
Version: AVK 16.5307 from 04.02.2006
|
