In message <[EMAIL PROTECTED]> Mike Smith writes:
: This is the usual poorly thought out argument, which fails to note that
: when you lose a disk you're already screwed due by /etc/fstab and the
: need to hard-mount local filesystems.
No. You aren't screwed. I have a system that needs /, /var, /usr and
/usr/local (all on one disk) to boot, but puts the fsck/mount of all
the other file systems into the background. If one of them fails, all
the others are still available.
: The "right" solution is and has always been to name your disks and mount
: them by name. Once devfs is a reality, we'll be able to do just this.
: Until then, the problem's not really as bad as you make it out to be.
I'd have to disagree with this. I was constantly getting burned by
the scsi system until I started hard wiring my scsi disks. The JAZ
drive I have had a period in its life when sometimes it wouldn't power
on (due to a bad power connection it was later discovered), so
sometimes at boot the system would see it and sometimes not. My /jaz
partition would wound up being where my /big partition should be and
my /big partition wouldn't be there at all when this happened. After
hardwiring, /jaz was the only one affected.
The same issue is there with ethernet interfaces too, btw. I have a
server that has 2 single fxp cards and 4 dual fxp cards. We need to
replace one of the single fxp cards with a dual one (no more slots and
need another interface), but to do this will require lots of
renumbering/cable shuffle, etc because ALL the interface numbers will
change when we do this.
I'm not convinced that the "right" solution is really "right" here.
Warner
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
