SunOS 4.1 - index on tape

dusan at batserver.cs.uq.oz.au dusan at batserver.cs.uq.oz.au
Fri Jan 25 10:34:37 AEST 1991


Some time ago I posted a message about the problems to put the index
(table of contents) at the beginning of the tape. I have encountered many
problems with SunOS 4.1. Actually I have an Exabyte drive EXB-8200
connected to a SparcStation 1+.

Well, I did not know that it was so complicated but I have the solution
which works for me and the only thing that makes me worried is: I do not
know the full explanation. Anyway, I would like to explain what I have
been doing for some time now.

Basically, here is what I am doing:

 1. mt -f /dev/nrst1 rewind
 2. dd if=/dev/rsd0d of=/dev/nrst1 conv=block,sync bs=56k count=1600
    (this initializes a big file at the beginning of a tape)
 3. mt -f /dev/nrst1 weof 2    (idea given by Chris Elvin) 
 3. dump (as many as I need file systems)
 4. mt -f /dev/nrst1 rew
 5. mt -f /dev/nrst1 fsf 3     
 6. restore -vft /dev/nrst1 >$File (done with a loop - to check
				    all file systems...)
 7. mt -f /dev/nrst1 rew
 8. tar cf /dev/nrst1 $File   (put the index on the tape...)


Now, I have tested this method using different levels of dumps (0, 5, 7,
3,...) which means that table of contents is different all the time... And
when I put them on the beginning of the tape - the method which I have to
use to reach the first dump after the index is

    MT -F /DEV/NRST1 FSF 3       (try this if you do not believe!)

It works fine!

Again, I have tried it about 30-40 times with different parameters and
dump levels and it is OK!!! (although the size of the index at the
beginning of the tape is never the same!)

I am puzzled and I do not know the explanation why it works like this (I
mean it is a bit strange) - BUT IT WORKS FOR ME.  Maybe someone has a
better solution.  I am looking forward to any comments and explanations. 

I would like to express my gratitude to the following people that have
sent very useful comments about Exabyte on the Net (and some of them to me
directly). Thanks for their effort and patience.

brad at athena.arc.nasa.gov (Margaret Bradshaw)
romig at cis.ohio-state.edu (Steve Romig)
aronb at gkcl.ists.ca (Aaron Burns)
fpb at ittc.wec.com (Frank P. Bresz)
C.Elvin at EE.Surrey.Ac.UK (Chris Elvin)
ghg at purdue.edu (George Goble)
msd at bo.ind.trw.com (Marc S. Dye)
tony at rata.vuw.ac.nz (Tony Martindale)
mlandau at bbn.com (Matt Landau)
felix at warbucks.ai.sri.com (Felix Ingrand)
zwicky at erg.sri.com (Elizabeth Zwicky)
marty at radarsun.Jpl.Nasa.Gov (Marty Slade)
jackal at epps.kodak.com (Philip Hammar)

Best wishes to everybody,

Dusan U. Baljevic
The Department of Computer Science
The University of Queensland
4072 - Australia
ph. (07) 365 3239
Amateur Radio VK4FCW
E-mail: dusan at batserver.cs.uq.oz.au



More information about the Comp.sys.sun mailing list