RE: [aclug-L] Funny problem with /etc/fstab

["Bates, Rod" <Rod.Bates@xxxxxxxxxxxxxxxxxx>]

Well, I found my immediate problem, but many questions
remain. unanswered  

By systematically commenting out lines in fstab, I 
discovered a missing comma between the options
on the cdrom line.  Fixing this seems to have made
the "line 0 ... is bad" symptom go away and stay.  

This error first appeared in fstab.~8~, which is dated
Jan 18.  I know I have done mounts many times since
then with no problem.   Maybe the upgraded version
of Linux is less tolerant of this error?  

But why the many inconsistencies even now in when
the symptom appears?  For example, I first commented
out all the nfs entries in fstab, and the message went
away, (even tho' the later-to-be-discovered bad entry
was still there), but reappeared after later experiments.

Why did /usr sometimes get mounted, sometimes not
during bootup?  Why did "df" and "mount" sometimes
(but not always) fail to list file systems that I could
navigate into?  

What are the entries with process ids in them?  They
appeared one at a time in versions with dates ranging
from oct 97 to apr 98.  I certainly didn't put them there,
nobody else has had access to do it, and the man page
for fstab says no programs write fstab.  They were not
commented out when my problems first began.  I did
that as one of the first things I tried. Now uncommenting
them does not cause any immediately visible problem. 

And finally, why have the nfs mounts started to silently
not happen? (I haven't mentioned this before).  They were
working immediately before the upgrade.  Now I can't
mount manually, using just the mount point,
(it claims the mount point is not found in either
/etc/mtab or /etc/fstab,)  but I can ping the servers
and mount manually using the full mount parameter list.