Re: [trill] TRILL Resilient Distribution Trees, who have read it?

[Mingui Zhang <[e-mail]>]

Hi Thomas,

Thanks for posting the question.

TRILL provides multicast service using ISIS link state routing. When there is a failure, RBridges are notified through the propagation of LSPs, which will trigger a campus-wide convergence on distribution trees. The convergence includes the following major processes: 1) the propagation of LSPs around the whole campus; 2) the recalculation of trees on affected RBridges; 3) the update of forwarding information on those RBridges. These processes make the convergence time-consuming. As we know, most multicast traffic is generated by applications which are sensitive to interrupt latency. Therefore, the convergence will lead to the disruption of multicast service.

Due to the propagation delay of LSPs, the distribution trees are calculated and installed on RBridges at different time, which brings _inconsistence_ of forwarding states among RBridges. Before the convergence finishes, multicast frames may be trapped by forwarding loops and finally dropped. This increases the volume of broadcast/multicast traffic. 

Protection is not a new topic for link state routing. It was proposed to reduce traffic disruption and resolve inconsistence during convergence using backup paths pre-installed. TRILL specified to calculate multiple distribution trees in those first days, which offers kind of ability for protection. The draft "TRILL Resilient Distribution Trees" makes use of this feature to provide the multicast protection without significant pay, which is natural rather than premature.

Thanks,
Mingui


>-----Original Message-----
>From: Thomas Narten [mailto:[e-mail]]
>Sent: Saturday, December 15, 2012 1:00 AM
>To: Mingui Zhang
>Cc: [e-mail]
>Subject: Re: [trill] TRILL Resilient Distribution Trees, who have read it?
>
>Hi.
>
>Per the Atlanta minutes:
>
>> Donald Eastlake (Huawei, co-Chair): OK how many people have read this
>> draft? That's a pretty small number so I suggest a message to the list
>> urging people to read this draft.
>>
>> Thomas Narten (IBM): I have a high level question. How important is
>> this document to the core of what we have to do as a WG? Seems like
>> this assumes you have a distribution tree, which we have now, and the
>> model is that if something goes wrong you just re-build the
>> tree. Maybe there is a brief period during which things don't
>> work. What we are doing here is building up a stand-by path. How
>> important is this? Do we know that convergence is too slow? It seems
>> to be to be pre-mature to work on this.
>>
>> Donald: I agree that that is the relevant question. How important is
>> this draft? Perhaps you should post that question to the list.
>
>I'm asking this question on the list.
>
>Thomas