[rbridge] loops in trill networks

[Gray, Eric]

Larry,

	During a bridge-level topology change, bridges only 
deal with BPDUs.  Other traffic is held up until the ST is
determined.  Otherwise, traffic in a bridged network would
be out of control and stay out of control.

--
Eric

--> -----Original Message-----
--> From: rbridge-bounces at postel.org 
--> [mailto:rbridge-bounces at postel.org]On
--> Behalf Of Larry Kreeger (kreeger)
--> Sent: Tuesday, October 11, 2005 4:45 PM
--> To: Developing a hybrid router/bridge.
--> Subject: Re: [rbridge] loops in trill networks
--> 
--> 
--> Joe,
--> 
--> The context was that during a topology change, there could be a
--> temporary loop.  While this temporary loop exists, 
--> multicast/broadcasts
--> would duplicate to edge ports each time they came around - 
--> until the TTL
--> reaches its limit.
--> 
-->  - Larry
--> 
--> -----Original Message-----
--> From: rbridge-bounces at postel.org 
--> [mailto:rbridge-bounces at postel.org] On
--> Behalf Of Joe Touch
--> Sent: Tuesday, October 11, 2005 1:02 PM
--> To: Developing a hybrid router/bridge.
--> Subject: Re: [rbridge] loops in trill networks
--> 
--> -----BEGIN PGP SIGNED MESSAGE-----
--> Hash: SHA1
--> 
--> 
--> 
--> Larry Kreeger (kreeger) wrote:
--> > OK, I can see how a TTL will stop loops, and would 
--> prevent duplicate 
--> > known unicasts from being delivered, but how does a TTL stop 
--> > broadcast/multicast/unknown unicast frames from being 
--> duplicated to 
--> > edge ports each time the frame comes round the loop?
--> 
--> The multicast/broadcast packets are forwarded by one or 
--> more spanning
--> trees. There should never be loops in those anyway as a result.
--> 
--> Joe
--> 
--> > Or is this considered
--> > OK behavior in an RBridge network compared to an 802.1 bridged
--> network?
--> > 
--> > New to all this...  - Larry
--> > 
--> > -----Original Message-----
--> > From: rbridge-bounces at postel.org 
[mailto:rbridge-bounces at postel.org] 
> On Behalf Of Joe Touch
> Sent: Tuesday, October 11, 2005 7:07 AM
> To: Developing a hybrid router/bridge.
> Subject: Re: [rbridge] loops in trill networks
> 
> There is, and that's what it's there for.
> 
> Joe
> 
> Mike Hughes wrote:
> 
>>--On 11 October 2005 15:23 +0200 Loa Andersson <loa at pi.se> wrote:
>>
>>
>>
>>>this might be an old discussion, but it is some time since it look at

>>>it. What do we do in trill networks about transient loops, i.e. if 
>>>use
> 
> 
>>>link state protocols there is a possibility that we have transient 
>>>loops. With no TTL mechanisms those looping frames could cause quite 
>>>a
> 
> 
>>>bit of problem.
>>
>>
>>Unless I've missed something, or things have changed overnight, there 
>>is a TTL provided in the SHIM header.
>>
>>Mike
> 
> _______________________________________________
> rbridge mailing list
> rbridge at postel.org
> http://www.postel.org/mailman/listinfo/rbridge
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (MingW32)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFDTBpUE5f5cImnZrsRAk2EAKDUBLgyz1duwtVTZyrUE8t++Z381wCdFW4L
OG6Pao5gsla9YrMI+qEdiNA=
=dTLU
-----END PGP SIGNATURE-----
_______________________________________________
rbridge mailing list
rbridge at postel.org
http://www.postel.org/mailman/listinfo/rbridge
_______________________________________________
rbridge mailing list
rbridge at postel.org
http://www.postel.org/mailman/listinfo/rbridge