| Technically speaking, there should not be any problems with a
PCM system monitoring itself. Just connect the console to a
terminal server and set it up as usual.
However, there are plenty of reasons not to do this. First and
foremost, the inevitable will take place sooner or later and
this inevitable is:
Someone is using PCM. Decides e.q. to shutdown a system, in this
unfortunately the PCM system.
Now, how does (s)he recover? Assume that (s)he might be e.q.
100km away using the PCM system over a mobile connection.
If there is need for PCM management of a PCM system, I'd really
go for TWO PCM systems each managing each other.
_veli
|
| -1 said it perfectly. I have always advised people not to do this. The
solution, as -1 stated, is to have another node running PCM. You can
also have the two PCM engines monitoring each other.
Now, there is nothing wrong with having the PCM engine "monitoring"
itself using one of the local async ports. The difference is this port
is just a terminal port and not the actual console port. You could
direct most software generated "console" messages to this port.
REgs,
Dan
|