3.51a keeps dying; even HDB is a mess; any ideas out there?

[Ed Horch]

Well, I'm running 3.5 (not 3.51[a]), and I experience a similar
problem with the system locking up for several seconds at a time
during heavy UUCP transfers via the OBM.  However, this seems to be
harmless, if a little annoying.  I run a full newsfeed here (yes, at
1200 baud; I'm saving my pennies for a Trailblazer), and have never
seen any harm done by this lockup problem.  It also doesn't generate
any messages in the unix.log file.

It does seem to grab the *entire* system, though, since the curser
even stops flashing.  I can't believe that something that can grab the
CPU for five to ten seconds at a time isn't a bug.

Speaking of bugs....

Can somebody send me the fix to rn to get around the "Can't open 
/dev/tty" message when dialing in from the OBM?

Also, is there any fix or workaround to the famous lost-inode problem
that occurs most often during news unbatching?  It's a real pain, more
because of its seemingly complete randomness than anything else.

Also, a reliability update:  I have been running continuously for over
seven months with exactly ZERO unexplainable crashes.  I have a 7300
(two fans) with a 1MB motherboard, and a Miniscribe 6085 disk that
sits outside the machine.

-Ed