>> For anyone else following this, under 'show global' you should see a line
>> like:
>> Assigned Address: 208.198.63.0 (Pool Size 32)
>
>Question: I've always used set assigned 208.198.63.1 (e.g. started with
>the first *usable* IP address in a subnet...not the subnet "boundary" IP).
>My reading of the Command Line Administrators Guide also leads me to
>believe that I should use 208.198.63.1
>
>And from the ComOS 3.5 Release Notes:
>
> The PortMaster allocates a pool of IP addresses *starting at the
> Assigned Address base value* (set from the global menu or by the
> "set assigned" command) and counting up. (emphasis added)
>[SNIP]
> In Release 3.5 the size of the pool can also be set explicitly
> with the "set pool Number" command, where Number will be the
> number of IP addresses allocated to the pool.
>
>So, *my* interpretation of the above is that if I set assigned to
>208.198.63.0 with a pool-size of 32, then the Portmaster will assign
>ALL the IP's from 208.198.63.0 - 208.198.63.31 inclusive. Wouldn't this
>cause a routing problem? (if not for OSPF, then for RIP?)
I've been hoping to see someone else answer this because I'm currently
doing this and I'm not sure if it is causing problems or not.
I've just been scanning through my radius logs to see what happens when
people get assigned an IP address on a subnet boundary. I've noticed that
about 3 out of 4 connections that get one of these addresses was dropped in
less than 4 minutes (maybe just long enough for the customer to be sure his
connection is not working???) . But a few times the connection was in use
for more than 30 minutes.
Could this be a problem for some user-end TCP/IP stacks but not others??
Anyone out there have a sound answer rather than my pure guesses??
Thanks,
- Randy Moore
Atlantech Online, Inc.
(301) 593-2794
http://www.atlantech.net/