I have the same problem. I called it in a few months ago to
Livingston, and worked with them for some time before Brian came out
on the list and said that it was a known bug and that Livingston was
unwilling to fix it.
However, I find it interesting that this ONLY happens with the
Ascend Pipeline. I have two customers that are having this problem.
They are our only ascend pipeline customers. Livingston showed me
how to pull teeth on the portmaster to stop it from speaking the
antiquated and useless RIP (you have to do this for every user, port,
and interface separtely. There ios no global "don't do it" command).
Ascend claims that the Pipelines that my customers are using don't
even have RIP programmed in.
So how is it that the portmaster can deduce that it is an Ascend
calling and drop the route for it? I find a workaround for the problem
considerably less tasteful than actually FIXING the problem. After all,
we are talking about removing a route from the routing table. This is
nontrivial.
It is no surprise to me that Livingston is more than 4 months overdue
on OSPF: they haven't even got static routing working, yet, and they are
still using the useless netmask table.
--
"There is an old readme file which says Jack McKinney
you should keep your friends close, jackmc@realtime.net
and your enemies even closer." Real/Time Communications (451-0046)
-Phong, Reboot Programmer/System Administrator