🚀 go-pugleaf

In article <1081@rsp.UUCP>, tom@rsp.UUCP (Thomas Ruf) writes:
|> macomber@thoreau.nsc.com ( Robert Macomber) writes:
|> 
|> >It seems to me that this argument assumes that the ailing station
|> >is still capable of responding intelligently to a trace signal.
|> No, my assumption is that the CPU of the system in question is so dead
|> that is does not react to anything (e.g. an OS Kernel that just panic()'ed).
|> After the "global" trace, all good guys will run a path test and then
|> restart PCM. Even the "smart" PCM chips like the DEC/AMD/Moto PLC
|> will not cycle through PCM signalling without CPU intervention.
|> As a consequence, the neighbor's CFM (ahm, sorry CEM ) would isolate
|> the broken machine, so the customer has enough time and reason to
|> blame the vendor ..
|> -- 
|> Thomas Ruf	tom@rsp.de	Schneider & Koch GmbH
|> {uunet,mcvax}!unido!rsp!tom	Germany

In a multiprocessor environment, where one processor controls the PCM
for one or more ports and a second processor controls the MAC and the
station's configurations, it is possible for the MAC's processor to
stop while PCM is still able to operate. Hopefully, the designer of
such a system will have some mechanism in place to detect either the
stopped processor or lack of communication between processors.

-- 
/======================================================================\
| Lawrence J. Pollack                   |                              |
| XLNT Designs, Inc.                    | Voice   : 619-487-9320       |
| Internet: larry@xlnt.com              | Fax     : 619-487-9768       |
\======================================================================/