I've got a PM-2E-30 with assigned pool of IP's starting at xxx.xxx.40.20
(top end's xxx.xxx.40.50). I've got a "dedicated" dialup on port s29
(phone #'s not in rotary pool) who's subnetted IP is xxx.xxx.41.0 with
active workstations on 41.1 thru 41.19 (running a telnet site/LAN). I've
got a permanent route set in the PM pointing xxx.xxx.41.0 at
xxx.xxx.40.50. Radius has the user's Framed-Address set to xxx.xxx.40.50.
so every time he logs in, his subnet works fine (ping to and from any IP,
telnet into his server, etc.).
This all works as advertised as long as the dedicated user's online.
However, if he's not, and the rest of the PM ports fill up, then the PM
assigns xxx.xxx.40.50 to any user when .50 becomes available from the IP
pool.
Then when the "dedicated" user logs in, he can't get his assigned IP of
xxx.xxx.20.50, since it's already in use. And since I've got a route
pointer to his subnet thru the .50 IP, incoming traffic tries to go to the
hapless user who's been assigned the IP for his session. Dismal failure.
How do I keep this from happening? The dedicated user's paying good
money for his connection, and the way it stands, I can't guarantee it.
What's the answer here? Did I skip school the day they taught this
in Portmaster 101?? A shiny apple to the good boy or girl who can
answer. Hell, it the answer's good enough (bulletproof), a bottle of
wine :-)
Any takers?
John Jamerson
Systems Administrator/WAN Manager
CoastalNet/Global Information Exchange
jamerson@coastalnet.com