From: Curtis Call (curtiscall@xxxxxxxx)
Date: Mon May 21 2001 - 04:28:01 GMT-3
I'm aware of that since it is standard split-horizon behavior. What
surprised me though was that SAPs are also suppressed on interfaces where
they are not learned if that interface is actually the shortest path to get
to the source network. I set up a lab that was circular in shape with four
routers. The top right router was advertising the static SAP, when it
reached the bottom right router it would keep it in it's SAPs table,
however it would not advertise it back to the top right router (split
horizon) nor would it advertise it to the bottom left router. What I
discovered was that a show ipx route showed that the path to the source
network of that particular SAP was through the bottom left network, so it
is obvious to me that Cisco's do an additional check to split horizon where
they will not send a SAP back towards the source.
At 12:07 PM 5/18/01, you wrote:
>Curtis: SAPs will not be broadcast out the interface they are learned
>on. As an example, configure static SAPs and, with default routes,
>point the static SAP network to different interfaces on the router. You
>will see the static SAPs advertised out the IPX enabled interfaces
>except the interface in the default route.
>
>Example: you have two Ethernet interfaces e0 with IPX network 10 and e1
>with IPX network 20. Define two static SAPs:
> ipx sap 04 Zero aaa.0.0.1 451 2
> ipx sap 04 One bbb.0.0.1 451 2
>Define two default routes, ipx route aaa 10.0000.1234.5678
> ipx route bbb 20.0000.5678.9abc
>You will see Zero advertised on e1 and One advertised on e 0.
>
>HTH, Fred.
>
>Curtis Call wrote:
> >
> > I noticed something that I don't remember seeing mentioned anywhere in the
> > documentation. Am I correct that IPX SAP propagation flows away from the
> > supposed sender? I assumed that a router treated RIP and SAP the same, it
> > just broadcasted all it's known routes and SAPS at set times out all RIP
> > enabled interfaces (keeping split horizon rules in mind) but in the lab I
> > am working on right now I've noticed that the router will suppress SAP
> > updates that flow towards the source, according to the routing table.
> > Is this standard behavior or am I just seeing things? I guess it follows
> > the same behavior as type-20 propagation, correct?
> > **Please read:http://www.groupstudy.com/list/posting.html
**Please read:http://www.groupstudy.com/list/posting.html
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:30:47 GMT-3