Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!ucbvax!CIVILGATE.CE.UIUC.EDU!lray
From: lray@CIVILGATE.CE.UIUC.EDU (Ray)
Newsgroups: comp.sys.apollo
Subject: Why is /etc/reboot so unreliable?
Message-ID: <8909291722.AA00609@civilgate.ce.uiuc.edu>
Date: 29 Sep 89 17:22:11 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 33


I've been having considerable problems with /etc/reboot. Over 50% of the
time, the node being rebooted fails to come up with one of the 
following symptoms:

   1. The node says the clock is off and waits forever for a y/n.
   2. The node freezes prior to going into self tests.
   3. The node fails to shut down at all.

Condition 1 is the most understandable. If there has been a crash and
the hardware clock has been scrambled, then it seems clear a sanity
check is in order.

Condition 2 could be something to do with the ROM level of the cpu.
I have not done enough testing to verify when this happens.

Condition 3. is the strangest. It acts like the node is in service mode
when I know for a fact it is in normal mode (yes, pressing the reset 
button works as expected).

The problem with all this is that I simply cannot easily get access to
the offices where there is a node at all times. This means the node
stays down for an unacceptably long period of time while I wait for
a key to become available.

---------------------------------------------------------------------

Does anyone know how /etc/reboot works, enough to give me some insight
as to whats going on here?

                                            Leland Ray
                                            UIUC - CE Dept.
                                            (217) 333-3821