SUMMARY: what is the real sense of those ufsdump-switches?

From: Bernt Christandl (beb@MPA-Garching.MPG.DE)
Date: Wed Jan 12 2000 - 06:53:55 CST


Hello managers,

i asked:
>Is there another sense in these "d", "s" and "b" switches for the
>ufsdump-command than to calculate how much time a specific dump will take?

and soon I did a mistake (asking for the "b"):
I knew that the "b" switch (blocking factor) is *very* useful to
optimize the speed of a dump to a specific kind of tape-drive:

Example from "David Evans" <DJEVANS@au.oracle.com>:
>On a DLT I changed blocking from default to
>65536 under cpio and dropped that backups from 12 hours to 3 hours. But
>you need to test as it can increase times as well

"d" and "s" - now i've learned this:
-> the old SunOS *required* these switches!
-> Solaris on the other hand is able to "sense" the end of a tape,
   if the drive is "modern enough"

This "modern enough" is obviously true for "my" DDS-[2/3] DAT-drives
and so i could (and can) safely ignore these switches.

I've been told, that "modern enough" is not true for those old 9-track
reel-tapes (are they still in use somewhere???), which still need these
switches.

One "vote" was, that multi volume dumps may need these too.
I never tried and so, i can't tell...

Thanx again to all of you who took time to answer me!

With regards,

Bernt Christandl

-- 
*****   Max-Planck-Institut fuer Astrophysik   *****
Bernt Christandl           Tel: +49 / 89 / 3299-3232
Postfach 1523              Fax: +49 / 89 / 3299-3235
Karl-Schwarzschildstrasse 1           85740 Garching
********   Email: beb@mpa-garching.mpg.de   ********



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:14:01 CDT