Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 17 Apr 1996 16:10:01 -0600 (MDT)
From:      Brandon Gillespie <brandon@tombstone.sunrem.com>
To:        questions@freebsd.org
Subject:   Adding a new disk, odd situation...
Message-ID:  <Pine.BSF.3.91.960417155224.4624A-100000@tombstone.sunrem.com>

next in thread | raw e-mail | index | archive | help
I have a server running FreeBSD 2.1-R with a kernel rebuild for the BOCA 
board.  When we upgraded the machine from a Linux system we swapped a few 
drives around, ending up with two SCSI drives.  The first was a new 
Conner 1.0GB SCSI F&W drive, the second was an older conner 1.3GB SCSI-2 
drive.  We were having problems with our SCSI controller on handling the 
first drive, so we opted to just ignore it (as we doubted the controller 
could handle it--odd situation, the drive actually has a SCSI-2 adaptor 
strapped on the back, and was intended for macs--which it works on just 
fine).  Instead we installed on the second drive, which reported itself 
on scsi id #6, although we dont know why (this drive has a history of its 
own, it came out of an AT&T star server).  We didn't want it on SCSI id 
#6, but we also could not find the appropriate jumpers to change it.  
Consequently we simply installed it as it stood.  We left both drives 
connected, with the first 1.0GB conner on scsi id #0, and installed to 
the second.  There are a few problems with this.  Primarily, to boot we 
have to boot with 'sd(1,a)/kernel' to get the right drive.

Act 2:

I now have two quantum SCSI drives I want to drop into the place of the
first drive.  One is a Quantum Lighting and the other is simply a Quantum 
(I dont recall the actual model, and I can't get to it at the moment).  
Replacing the conner with the quantum lighting resulted in some very 
interesting behavior, from the drive controller (before we even hit the 
OS); usually the controller (adaptech) reports each drive it finds upon 
bootup, similar to

    DRIVE-NAME   DRIVE-SPECS     #ID    C: (etc)

With the lighting in place it came up on id 0, with drive specs, then id 
1-5 were also the quantum lighting, but only in name (it didn't give the 
specs (size etc)), and finally ending with the conner 1.3 on id 6.  When 
I reached the boot prompt I intended to boot to the conner, but I 
couldn't find the filesystem device for it anymore... (it was no longer 
on sd0).  I could have booted from a floppy to let the OS probe the 
device, but I was rather rushed at the time, so instead I simply removed 
all drives other than the conner 1.3 gb and did another reboot, this time 
specifying sd(0,a)/kernel.  It reached the point of doing an fsck when it 
came up with a device-not-configured error (for obvious reasons, as sd1 
was the configured device)...  At this point I simply plugged back in the 
original conner to sd0 and rebooted to sd1 on the second conner without 
problems, leaving it for another day..

Act 3:

This saturday I have aquired more available down-time to work on the 
problem.  What I would like to know is what tools are available, and what 
would have to happen to get around specific possibilities, namely:

    1) How would one reconfigure a different scsi device for an existing
       filesystem?  This is in relation to if I end up having to keep
       the existing filesystem (which we need) on a device other than
       sd1...
    2) with the new disk, how do I make it a 'bootable' disk?  I.e.
       I doubt simply copying the core filesystem from the current disk
       to the new disk will work (once the new disk is formatted).  What
       I would really like is the possibility of using dd to just copy
       the entire disk across, except for the disk sizes are not the same
       (the quantum is only 540 MB), even though we have only used
       400MB of the 1.3 GB available.  I would like to set it up to boot
       to the quantum, with the core filesystem on the quantum and the
       users on the conner...

-Brandon Gillespie



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.91.960417155224.4624A-100000>