jw@sics.se (Johan Widen) (08/08/87)
What are the good and bad points of the QIC-11 and QIC-24 tape formats?
We have a Sun which supports both formats and I would like to know if one
of these formats is better in some sense.
I would also like to know how to calculate dump(8) tape length for a say
9 track cartidge tape. I do not have the source for Sun dump but I suspect
that dump regards all tapes as 8 track (8 data plus parity). My guess is also
that 9 track cartridge tape is used by writing 8 data bits and 1 parity bit.
Is this the case?
The formula I am using right now for calculating tape size is as follows:
I use the tape size written on the cassette with no safety margin!
density = 10000 (thats ten thousand)
size = 450 feet
My guess is that one of the following formulas is correct:
(tape size)*((9 tracks)/(8 tracks))*(safety factor)
(tape size)*(safety factor)
where (safety factor) is something like 0.9 or 0.94 (depending on who
you talk to).
--
Johan Widen
SICS, PO Box 1263, S-163 13 SPANGA, SWEDEN
{mcvax,munnari,cernvax,diku,inria,prlb2,penet,ukc,unido}!enea!sics.se!jw
Internet: jw@sics.seleres@ucbarpa.Berkeley.EDU (Craig Leres) (08/10/87)
This is a question I recently answered on comp.sys.sun. The QIC-24 format makes 9 passes over the tape. The general formula for figuring out how many feet of tape dump can use is: dump_tape_length = (actual_tape_length - 25) * no_of_passes So when doing dumps, you should use something similar to one of the following: dump 0uncbsf 126 2475 /dev/rst8 /mnt # QIC-24, 300 foot tape dump 0uncbsf 126 3825 /dev/rst8 /mnt # QIC-24, 450 foot tape dump 0uncbsf 126 5175 /dev/rst8 /mnt # QIC-24, 600 foot tape QIC-11 only makes 4 passes: dump 0uncbsf 126 1100 /dev/rst0 /mnt # QIC-11, 300 foot tape dump 0uncbsf 126 1700 /dev/rst0 /mnt # QIC-11, 450 foot tape dump 0uncbsf 126 2300 /dev/rst0 /mnt # QIC-11, 600 foot tape I think the recent consensus is that a block size of 126 is optimal. Craig