On Saturday 07 June 2003 08:06, Gene Heskett wrote:
>Greets everybody;
>
>Last nite I used a different, slightly older model of that Seagate
>CTL-96 drive, which seemed to work, with one glareing exception.  My
>little extra script does a 'tell' operaion on the tape before it
>tries to write the seperate indice and config files to the end of
> it, and it failed to write because the 'tell' returned a value many
> times the expected value for a 32k blocksize tape.
>
>TBE it returned a value indicating this tape was made with a
> blocksize of 32 bytes!
>
>But no other errors were reported.  So it looks as if I'd better
> cycle thru the other 3 tapes, (they are freshly labeled tapes since
> those 4 tapes are stuck in the other failed drive), doing a dd on
> the label and checking to see if they are at block 64 after reading
> the label.
>
>What is the std fix for this?  I'd done an "mt -f /dev/nst0 setblk
>32k" last nite.  Did mt not recognize the 'k'?

Poor form, replying to ones own posts, or so I've been told.  FWIW, 
the manpage on mt doesn't mention the suffux 'k' in the same 
paragraph as the setblk command.

OTOH, we've habitually used it for the dd command, so I'd assume 
(there's THAT word again) that dd must translate before it sends it 
off to the drive.

Anyway, it turns out that my std formatted tapes 'tell' returned value 
is based on a 512 byte block, so I am in the process of fixing the 
remaining tapes in the magazine to be 512 byte blocks now.

Can anyone discuss, from a bit of knowledge about such things, what 
the advantages and disadvantages are of a DDS2 tape at a block size 
of 512 bytes vs one with a block size of 32768 bytes?

Can I infer that the smaller block means a more frequent checksum and 
error recovery attempt?

At the cost of how much additional space overhead on the tape?

If not, how does this work in the real world?

-- 
Cheers, Gene
AMD K6-III@500mhz 320M
Athlon1600XP@1400mhz  512M
99.26% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com attornies please note, additions to this message
by Gene Heskett are:
Copyright 2003 by Maurice Eugene Heskett, all rights reserved.