Re: Two DLSW Questions.... posted 01/11/2001
On Tue, 9 Jan 2001, Robert DeVito wrote:
> 1.) When peering from a router with a ethernet interface accross the WAN to
> a router with a token ring interface, is it necessary to define the maximum
> fame size, i.e. "lf 1500" in the remote-peer statements? I have seen it done
> both ways and want to know the "correct" way.
if you use 'dlsw remote-peer lf 1500', it guarantees negotiation to 1500
bytes. It's necessary sometimes depending on the end stations, whether
they can actually handle segment sizes larger.
> e0 s0 s0 e0
> 2.) |---R1-----/-----R2---|
> When setting up DLSW from R1's e0 to R2's e0. I understand the local-peer
> address would be e0's ip address. But when I set up the remote-peer address
> statement, would the remote-peer statement on R1 be R2's E0 or R2's S0? I
> have seen it done both ways also.
The rule is: always use a router's local-peer id to configure for
remote-peer id in another router.
The deal is if you have r1's e0 address as the local-peer id, then that ip
address should appear as r2 remote-peer id. If r1 chose to use r2 s0 ip as
remote-peer id, then that ip should be r2's local-peer id.
It is the admin's choice to use which ip address as their local-peer
id. So some people just standardise on e0's ip address, for example, and
it's not a rule of thumb. In reality/production, one would use a loopback
interface, and configure an ip address there for peer id.
Deal there is that unlike e0, the interface can never be affected by any
physical issues like losing a carrier/keepalive, and goes down. if the
interface goes down, dlsw peer is down. Loopback interface doesn't suffer
from that syndrome. You have to ensure the ip is reachable.
Drawback is it generates host routes in their routing tables.. Plan
summarization is the key... :^)..
> Thanks in advance!
> Robert DeVitoRobert DeVito
> Get your FREE download of MSN Explorer at http://explorer.msn.com
> To unsubscribe from the CCIELAB list, send a message to
> majordomo@xxxxxxxxxxxxxx with the body containing:
> unsubscribe ccielab
To unsubscribe from the CCIELAB list, send a message to
majordomo@xxxxxxxxxxxxxx with the body containing: