🚀 go-pugleaf

[A copy of this has been sent to 'bug_next@next.com'.  I though it might
help to post another copy here...]

I just finished helping a another NeXT user deal with some problems
involving cron and mounting a second SCSI disk.  Now I realize that cron
needing execution paths to be happy is one of those little details that
users need to share & talk about (although a next-answer on it would be
nice), but the disk problem is a bit more critical.

It seems that his cron job (which did a cp of his current files over to
the backup (secondary) drive) would return an error message that it was
unhappy and could not find his drive, dispite the fact that he could see
it was mounted in his workspace when he logged in.  Of course, the real
problem was that his workspace was autodiskmounting his disk for him
whenever he logged in, and later when cron ran (& he was not logged in),
the disk was unavailable...  The solution:  Remove the 'noauto' option
from his /etc/fstab file line for the drive.

Now this is a fairly common problem, and I was all set to tell him to
RTFM (with the exact manual locations), when I realized YOU HAVE THIS
WRONG IN YOUR ONLINE SYSADMIN MANUALS!!!  Specifically, version 3.0,
/NextLibrary/Documentation/NextAdmin/07_Peripherals/01_SCSIDisks.rtfd
repeatedly states to use the noauto option in fstab (for the new disk
drive), which nicely voids that cute little statement 'mount -vat 4.3'
in the rc scripts...  The documentation then goes on to discribe the
behavior patterns of missing the noauto option in this /etc/fstab line.

[Note to others:  The noauto option is NECESSARY for the root partition,
but not desirable for other drives that you also wish to automatically
mount upon booting.  See mount(8).]

So, how about a next-answer on this?  Or a 'updated docs/release notes'?


Thanks,

    -David Apfelbaum.
     da0g+@andrew.cmu.edu