[101806] in RedHat Linux List
Re: diskcopy for linux
daemon@ATHENA.MIT.EDU (David E. Fox)
Sun Nov 29 15:48:21 1998
From: "David E. Fox" <dfox@belvdere.vip.best.com>
To: redhat-list@redhat.com
Date: Sun, 29 Nov 1998 12:45:44 -0800 (PST)
Reply-To: dfox@belvdere.vip.best.com
In-Reply-To: <19981129145151.A4791@fcshome.stoneham.ma.us> from "Fred Smith" at Nov 29, 98 02:51:51 pm
Resent-From: redhat-list@redhat.com
> Note that the 18432 used as the value of 'bs' is not critical, it could be
> almost anything, bigger is generally better. Best is if it is a whole-
> number multiple of the diskette track size (for 3.5HDdrives 18 sectors *
I haven't played that much with bs= values in dd, at least on Linux.
One thing I did notice early on when first using Linux is that having
a bs= value wasn't all that crucial, at least in comparison with
early versions of BSD (precursors to FreeBSD) I used previously. On
those, having a bs= value was absolutely necessary; otherwise, the
system would read a sector, write a sector, read a sector, write
a sector, and generally run *very* slowly. Just upping that to 10K or
so made an incredible difference back then.
I suspect it has something to do with buffering - afaik linux reserves
a whole disk track someplace in the kernel data (floppy track buffer) so
perhaps having a bs= is superfluous, at least when dealing with
floppies.
> Fred
------------------------------------------------------------------------
David E. Fox Tax Thanks for letting me
dfox@belvdere.vip.best.com the change magnetic patterns
root@belvedere.sbay.org churches on your hard disk.
-----------------------------------------------------------------------
--
PLEASE read the Red Hat FAQ, Tips, Errata and the MAILING LIST ARCHIVES!
http://www.redhat.com http://archive.redhat.com
To unsubscribe: mail redhat-list-request@redhat.com with
"unsubscribe" as the Subject.
