RE: DLSW direct encaps

From: Jason Sinclair (sinclairj@xxxxxxxxxxxxxxx)
Date: Thu Mar 14 2002 - 19:35:58 GMT-3


   
Hickito,

For direct encaps it is not the RIF that is passed-thru it is that the LLC2
frame is not locallly-acked and is acked via the remote end. Now when
talking about mapping the rules are as follows:

When using direct

Dlsw remote-peer 0 frame-relay interface serial 0 100 pass-thru ##### must
use frame-relay map dlsw 100

With lite

Dlsw remote-peer frame-relay interface serial 0 100 ######## must use
frame-relay map llc2 100

This is because when using light the data is reliably transported in LLC2.
With direct encaps it is transported via DLSW transport.

Let me know if you need more info.

Cheers,

Jason Sinclair
Manager, Network Support Group
POWERTEL
Ground Level, 55 Clarence Street,
SYDNEY NSW 2000
AUSTRALIA
office: + 61 2 8264 3820
mobile: + 61 416 105 858
* sinclairj@powertel.com.au

                -----Original Message-----
                From: Narvaez, Pablo [mailto:Pablo.Narvaez@getronics.com]
                Sent: Friday, 15 March 2002 04:30
                To: Jason Sinclair; A Yigit Zorlu; Bob Sinclair
                Cc: ccielab@groupstudy.com
                Subject: RE: DLSW direct encaps

                Hey Guys, I was looking at the posts and couldn't get a
clear idea of DLSw Direct encap y Lite encap ..

                I understand that Lite does local ack whereas Direct encap
let the RIF pass thru the link, right? ...

                Now, regarding the configuration please correct me if wrong:

                Direct encap:
                dlsw local-peer peer-id 1.1.1.1
                dlsw remote-peer 0 frame-relay interface serial 0 100
pass-thru

                int ser 0
                ip add 1.1.1.2 255.255.255.0
                frame-relay map llc2 100 brodcast

                Direct encap:
                dlsw local-peer peer-id 1.1.1.1
                dlsw remote-peer 0 frame-relay interface serial 0 100

                int ser 0
                ip add 1.1.1.2 255.255.255.0
                frame-relay map llc2 100 brodcast

                Is it that the only one difference between these two encap
modes is the keyword "pass-thru" in the end of the remote-peer
                statement? ...

                Aside from that, I have the option to use in the end of the
remote-peer the keyword "rif-pass-thru"? what is the difference
                between "pass-thru" and "rif-pass-thru"? which should I use
to accomplish Direct and Lite encap?

                And one more thing, in the interface I can use either:
                frame-relay map llc2 or frame-relay dlsw .... what is the
"fr map dlsw" command for? and when should I use it? ...

                Thanks in advance!!!

                Cheers,

                -hockito-

                -----Original Message-----
                From: Jason Sinclair [mailto:sinclairj@powertel.com.au]
                Sent: Lunes, 11 de Marzo de 2002 05:02 p.m.
                To: 'A Yigit Zorlu'; Jason Sinclair; 'Bob Sinclair'
                Cc: ccielab@groupstudy.com
                Subject: RE: DLSW direct encaps

                Yigit,

                You use it for the local-peer. Let's say you have an address
on your token
                ring, loopback and serial int. You must use the local-peer
peer-id with the
                address of the serial int NOT the loop or the token. I have
also verified
                this behaviour with Cisco and they have concurred with these
findings. I
                believe they are updating the doco appropriately.

                Regards,

                Jason Sinclair
                Manager, Network Support Group
                POWERTEL
                Ground Level, 55 Clarence Street,
                SYDNEY NSW 2000
                AUSTRALIA
                office: + 61 2 8264 3820
                mobile: + 61 416 105 858
                * sinclairj@powertel.com.au

                        -----Original Message-----
                        From: A Yigit Zorlu [mailto:alec_cisco@yahoo.com]
                        Sent: Monday, 11 March 2002 17:04
                        To: 'Jason Sinclair'; 'Bob Sinclair'
                        Cc: ccielab@groupstudy.com
                        Subject: RE: DLSW direct encaps

                        Jason,

                        I am confused. Where do you use ip address in direct
encapsulation ?

                        Please check :
                        o Dlsw local-peer
                        o dlsw remote-peer 0 frame-relay interface
serial 0 100
                pass-thru
                        o interface serial 0
                        o frame-relay map dlsw 100 br

                        Yigit

                        -----Original Message-----
                        From: nobody@groupstudy.com
[mailto:nobody@groupstudy.com]On Behalf
                Of
                        Jason Sinclair
                        Sent: Monday, March 11, 2002 3:15 AM
                        To: 'Bob Sinclair'; Jason Sinclair
                        Cc: ccielab@groupstudy.com
                        Subject: RE: DLSW direct encaps

                        Bob,

                        Direct must be point to point. When using direct you
must also use
                the IP
                        address on the serial int NOT a loopback or a
token/ether, etc. I am
                not
                        sure why, but from my testing the remote side must
ack the frames
                and thus
                        when you use any int that is not the directly
connected int, the L2
                frame is
                        then routed to the next int. That is also why only
p2p works I
                guess!
                        Remember with DLSW Lite this does not matter as the
frame is locally
                        acknowledged. This is only an issue when using the
pass-thru
                command.

                        I have some dynamic peer stuff that works. Basically
dynamic peer
                circuits
                        and the peer connection are only brought up when
there is traffic
                destined
                        for the remote side (I guess that is fairly
obvious). I have had
                success
                        doing this when using a dest-mac statement under the
remote-peer as
                you can
                        better control the dynamic peer. Some configs are as
follows which
                create a
                        host and a FEP and also two DLSW routers runnning
dynamic over HDLC.
                Nothing
                        fancy, I am not using loopbacks just the serial
ints.

                        When you start it up the peers will connect and 2
circuits will
                establish.
                        If you shut the token on HostA, the circuits will
drop and after 20
                minutes
                        the peers will disconnect. You can drop this timer
if that is too
                long. Also
                        as I have specified the dest-mac, the peers will
only establish for
                this
                        address as dynamic peers only establish after all
filter rules are
                met.

                        Also remember to bit-swap the mac addresses on the
host or FEP if
                you are
                        doing token to ether as you are statically defining
the remote mac
                and dlsw
                        will not swap it for you in this instance.

                        Let me know if you have any trouble using this or
have other DLSW
                questions.
                        I am at work right now and can't access my home lab,
but if you have
                any
                        problems I will re-create tonight. This is all from
the top of my
                head at
                        the moment!!

                        
        
HostA-------tokenring---------DLSWA--------serial--------DLSWB--------tokenr
                        ing----------FEP

                        HostA (PU2)
                        dspu host PU2 xid-snd 01712345 rmac 4000.1111.0001
rsap 4 lsap 4
                        retry-timeout 5
                        !
                        dspu host PU3 xid-snd 01712345 rmac 4000.1111.0001
rsap 8 lsap 8
                        retry-timeout 5
                        !
                        interface TokenRing0
                         mac-address 4000.3000.0002
                         no ip address
                         ring-speed 16
                         dspu enable-host lsap 4
                         dspu enable-host lsap 8
                         dspu start PU2
                         dspu start PU3

                        DLSWA
                        Dlsw local-peer peer-id 1.1.1.1
                        Dlsw remote-peer 0 tcp 1.1.1.2 dynamic inactivity 20
dest-mac
                4000.1111.0001

                        !
                        Int s0
                        Ip address 1.1.1.1 255.255.255.252

                        DLSWB
                        Dlsw local-peer peer-id 1.1.1.2 promiscuous
                        !
                        int s0
                        ip address 1.1.1.2 255.255.255.252
                        clock rate 64000

                        FEP(PU4)
                        dspu pu PU2 xid-rcv 01712345
                        !
                        dspu pu PU3 xid-rcv 01712345
                        !
                        interface TokenRing0
                         mac-address 4000.1111.0001
                         no ip address
                         ring-speed 16
                         dspu enable-pu lsap 4
                         dspu enable-pu lsap 8

                        Cheers,

                        Jason Sinclair
                        Manager, Network Support Group
                        POWERTEL
                        Ground Level, 55 Clarence Street,
                        SYDNEY NSW 2000
                        AUSTRALIA
                        office: + 61 2 8264 3820
                        mobile: + 61 416 105 858
                        * sinclairj@powertel.com.au

                                        -----Original Message-----
                                        From: Bob Sinclair
[mailto:bsin@erols.com]
                                        Sent: Monday, 11 March 2002 09:03
                                        To: Jason Sinclair
                                        Cc: ccielab@groupstudy.com
                                        Subject: Re: DLSW direct
encaps

                                        Jason,

                                        Since you have some experience with
this, maybe you
                could
                        confirm my understanding. Is it true that direct
encapsulation over
                frame
                        relay is only point-point. I don't mean subif type,
but that if you
                had a
                        hub and spoke configuration, one could not do direct
encap spoke to
                spoke.
                        Does that sound right?

                                        Also, do you have a working config
for dynamic
                peers? I
                        have not been able to get this to work or to find a
good example.

                                        Thanks

                                        ----- Original Message -----
                                        From: "Jason Sinclair"
<sinclairj@powertel.com.au>
                                        To: <ccielab@groupstudy.com>
                                        Sent: Sunday, March 10, 2002 6:24 PM
                                        Subject: DLSW direct encaps

> All,
>
> FYI - direct encapsulation over
frame relay with
                IETF does
                        not work in any
> version of IOS from 12.0 - 12.2
(don't know about
                before).
                        It works fine
> with Cisco encaps however. I spent
three days
                playing with
                        this and in the
> end using a frame analyser and
debug frame packet
                you see
                        that DLSW
> manipulates the IETF frame to the
point where it
                has an
                        "ILLEGA:" value in
> it. Just thought this might be
interesting to
                someone.
>
> Regards,
>
> PS - I have the debug output if
anyone wants it.
>
> Jason Sinclair
> Manager, Network Support Group
> POWERTEL
> Ground Level, 55 Clarence Street,
> SYDNEY NSW 2000
> AUSTRALIA
> office: + 61 2 8264 3820
> mobile: + 61 416 105 858
> * sinclairj@powertel.com.au
>
>
>
>
>
                        
        
**********************************************************************
> PowerTel Limited, winners of
> Broadband Wholesale Carrier of the
year,
                CommsWorld
                        Telecomms Awards 2001
> Best Emerging Telco, Australian
Telecom Awards
                2001
>
>
                        
        
**********************************************************************
> This email (including all
attachments) is intended
                solely
                        for the named
> addressee. It is confidential and
may contain
                commercially
                        sensitive
> information. If you receive it in
error, please
                let us
                        know by reply email,
> delete it from your system and
destroy any copies.
>
> This email is also subject to
copyright. No part
                of it
                        should be reproduced,
> adapted or transmitted without the
prior written
                consent
                        of the copyright owner.
>
> Emails may be interfered with, may
contain
                computer
                        viruses or other defects
> and may not be successfully
replicated on other
                systems.
                        We give no
> warranties in relation to these
matters. If you
                have any
                        doubts about
> the authenticity of an email
purportedly sent by
                us,
                        please contact us
> immediately.
>
>
                        
        
**********************************************************************
>
        



This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:57:09 GMT-3