Date: Mon, 24 Jun 1996 13:12:16 +0200 (MET DST) From: "Georg-W. Koltermann" <gwk@racer.dkrz.de> To: abladtn@autelca.ascom.ch Cc: gilles.tschopp@aletsch.esis.vsnet.ch, scsi@FreeBSD.org Subject: Re: How to use Syquest EZ-135 as backup drive ? Message-ID: <199606241112.NAA11225@racer.dkrz.de> In-Reply-To: <31CE34F6.794BDF32@autelca.ascom.ch> (message from Norbert Bladt on Mon, 24 Jun 1996 08:25:58 %2B0200)
next in thread | previous in thread | raw e-mail | index | archive | help
>>>>> "Norbert" == Norbert Bladt <abladtn@autelca.ascom.ch> writes:
Norbert> Georg-W. Koltermann wrote:
>> >>>>> "Gilles" == Gilles Tschopp
>> <gilles.tschopp@aletsch.esis.vsnet.ch> writes:
Gilles> Hello, I'm an beginner of the Unix world. I would like to
Gilles> know how to manage backup of my FreeBSD system with the
Gilles> TAR command. I have a Syquest EZ-135 plugged on the SCSI
Gilles> subsystem. How to do ?
Gilles>
Gilles> A lot of thanks.
Gilles>
Gilles> Gilles Tschopp.
Gilles> ******************************************* Gilles
Gilles> Tschopp, Informaticien de gestion ES
Gilles> gilles.tschopp@esis.vsnet.ch http://icare.itech.ch
Gilles> http://esis.vsnet.ch http://www.vsnet.ch
Gilles> *******************************************
Gilles>
Gilles>
>> Try cd /someplace
Norbert> This is OK.
>> tar cvfb /dev/rst0 126 somefiles
Norbert> This is a bad idea and a wrong answer, too. I think that
Norbert> the EZ-135 is not a tape drive and therefore it is not
Norbert> detected as "rst0" ! Instead it will be some sort of
Norbert> harddisk.
Ouch!! Thank god it's not a scanner. I wouldn't know how to backup
to a scanner...
I realized when I sent out my first reply that there wasn't much
information from Gilles' original question. I only wanted to give him
a quick hint without going too much into details--he mentioned being a
newbie. I simply picked the most probable case, and hoped he would
come back to me if he got 'no such device or address'.
If this thing *is* a disk, and Gilles is a newbie, he'll probably have
a hard time to get it disklabel(8)'ed and newfs'ed. I think the
easiest route for him would then be to
a) Unplug any SCSI disks that might be in use, except for the
EZ-135. This is for safety, so that any subsequent
mistakes will not wipe out system files.
b) Start FreeBSD installation, selecting FLOPPY as
installation medium (only the boot disk will actually be
needed).
c) Select the EZ-135 as the target for the install. In the
FDISK menu create a FreeBSD slice spanning the whole disk.
In the disklabel menu create at least one dummy partition
for the root filesystem. That's the minimum to let the
install program continue, I believe.
d) When the install program continues, it will disklabel the
EZ-135. It will continue some more and ask for the root
floppy. At this point the installation can be aborted
(actually it could even be aborted right after the
disklabel step, but let's be sure and wait a little bit).
e) Reconnect the devices from a). Boot as usual, watching
the SCSI probe messages and take note which "sdX" the
EZ-135 is.
f) Check the /dev directory for an entry for the "sdX"
(substitute actual unit number for X) found during the
probe. Both the basic "sdX" and the "sdX{a,b,c...}"
partitions should be there. If they aren't, run "MAKDEDEV
sdXa" while in the /dev directory.
It should now be possible to access the EZ-135. The whole volume will
be /dev/sdXc. Any partitions created in c) would be usable also, if
one desires to subdivide the unit.
The remaining steps depend on the backup tools used: Tar could write
directly to the device or any partitions (the "raw" device or
partitions /dev/rsdXy would be preferred for tar). In order to use
cp(1), one would need to run newfs(8) on the device or partition
first, and then mount(8) it.
Georg-W. Koltermann, gwk@cray.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199606241112.NAA11225>