I've been given the responsibility of managing a deployment
of a TNT DSL installation for a small municipal-utility ISP.
I'm trying to figure out what service model we're going use.
Currently, we're going to be using DSLPipe CPEs for both SDSL and
ADSL (the ADSL uses CAP, not DMT, and I have no control over that).
As a user of DSL myself, I've been partial to bridging with static
IP adresses. I've deployed this setup at another ISP using a
Cisco and IRB (integrated routing and bridging), over Verizon's
older frame-relay based service. As you know, of course, the TNT
DSL supports bridging, PPP, frame-relay, etc, as do the DSLPipes.
We'd prefer to not be married to the DSLPipes, however, so we would
have the option of using other CPEs. The TNT manuals I have are
older (copyright 1998) and make no mention of bridging (nor do they
even refer to the OS as TAOS). What I want to know is:
1. Is bridging w/static IP addresses a workable model?
I'd prefer to use neither PPP nor other layer 2 protocols at
all (yes, I know a layer 2 protocol is necessary; I'm assuming
with bridging I'm just bridging Ethernet--I don't want the
clients to have to set up any special software other than
TCP/IP configuration in their client OSes).
2. Will it be easy with this model to assign additional IP
addresses, should the client want one?
3. Can I prevent users from just taking an unassigned IP address
and using it with this model? I know that with the frame and
IRB setup I'm using I associate the DLCI with the IP addresses,
but this isn't configured the same way.
4. Should I be able to use most other RADSL-CAP CPEs and which
ones do people have experience with and recommend?
5. Where can I get these cables with 50-pin amphenol
connector to 4x DB-37 connectors?
6. The guy providing the upstream bandwidth is claiming that
he can't give us IP addys to assign to every client, due to
"new ARIN rules". I've asked him for references, but not
gotten any. I think he's probably mistaken or full of it; does
anyone know anything about this? I'd prefer to not have to do
NAT or masquerading because it breaks some client applications
(not to mention complicating my dynamic routing), but I have
a Linux router if necessary.
Wil
--
Wil Cooley wcooley at nakedape.cc
Naked Ape Consulting http://nakedape.cc
* * * * Linux, UNIX, Networking and Security Solutions * * * *
QCSNet http://www.qcsn.com
* * * * T1, Frame Relay, DSL, Dial-up, and Web Hosting * * * *
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 232 bytes
Desc: not available
Url : http://shadowknight.real-time.com/pipermail/rte-ascend/attachments/20020820/63f50688/attachment.pgp