RE: [RPRWG] Questions about context containment

["Castellano, Robert" <RCastellano@xxxxxxxxx>]

Hi Walker,

 

It is true that all strict mode traffic is purged during context containment.  A station initiates context containment once it sees protection message on the ring (D2.4, clause 6.6.1.1). It is also true that the context containment period continues until the new topology has stabilized, and hence strict data frames will be purged until the context containment period completes.  This is to ensure that frames from the old context are completely removed before introducing new strict mode frames.  This is the only way to maintain ordering and prevent duplication. 

 

Also, there is no way to predict whether a particular transit frame will cross the edge.  If the frame in the old context is not purged before launching a new frame on the other ringlet, the edge could heal, allowing the frame from either old or new context to cross the edge, resulting in a misordering or duplication. 

 

            thanks,

 

            robert

-----Original Message-----
From: zhang xiaolin [mailto:bit_walker@xxxxxxxxxx]
Sent: Tuesday, September 09, 2003 5:22 AM
To: stds-802-17@xxxxxxxx
Subject: [RPRWG] Questions about context containment

Hi, everyone:

 

Provided that all the data frames transmited in the RPR ring are strict order and the protection mode is steer, will all the data frames be removed from the ring when a link failure occurs, in spite of whether they across the edge?

Only when the topology becomes stable and valid, context containment can be cleared. Is it true that all strict data frames, including the steered frames, can not arrive at the destination before context containment is cleared?

 

Thanks.

 

Walker