Assuming 4 nodes - A, B, C, and D, connected in SONET ring running UPSR. According to a Cisco press book, if a STS is mapped between two nodes (say A and B), then that STS is consumed for the entire ring and cannont be used to add/drop traffic elsewhere. However, the same book also mentioned that you can reuse the SAME STS between say, node C and D, if you leave them unprotected.
How does that work? If the STS is consumed by the A to B mapping, then it is consumed, period. How is that you can use that same STS to pass traffic between node C and D?
If you build the circuit _unprotected_ between A and B the same STS will be free everywhere else on the ring, HOWEVER, traffic is unprotected and if there is a fiber cut any traffic running between the nodes affected by the cut will drop.
If you use BLSR, then a conection between A and B would be protected by the the Sonet line. So if the circuit rides STS #1 between A and B and there is a fiber cut, the traffic wraps around and rides STS #25. The limitaton is that from any node there are 24 connection to the east and 24 connections to the west. However this is also an asset, in a 4 node ring, assuming all traffic was only to the adjacent node, there could be 96 circuits on the entire ring! (not likly in the real world).
Thanks for the reply, but I am still not quite clear..
1. So in a given STS, you can't mix and match protected and unprotected traffic, and it has to be either all protected or none unprotected?
2. Assuming a 2 fiber physical ring topology, when you provision the circuits unprotected, does that mean the circuit no longer has to obey rules such as UPSR? Does the traffic between node A and B still have to go clockwise, or does the concept of clockwise/counterclockwise no longer apply and it simply becomes a linear/point-to-point deployment?
The short answer is yes they can be mixed and matched.
If the ring is UPSR and the circuit is mapped between points A and B, the STS that is used between the 2 slots is used and cannot be used again. So lets say that node A uses slots 6 and 12 (in a 15454) and node B uses slot 6 and 12. The connection is between slot 12 of node A and Slot 6 of node B. (and continue this pattern around the ring) If STS-1 #1 is used from slot 12 of node A to slot 6 of node B, then STS-1 #1 cannot be reused between the same two points. However a different UNPROTECTED circuit could be built from slot 6 of Node A to Slot 12 of node D on STS-1 #1. So this is an example of reuse, however, normally in general practice you do not build circuits this way.
When you build a circuit on a UPSR usually a "path selector" is built at each of the drop nodes and the path selector determines which direction it is receiving from. In the transmit direction it is bridged in both directions. The path selector just has to figure out which receive it likes the best.
On your question 2, an unprotected circuit is just that, an unprotected circuit, it does not go around the ring at all! It is point-to-point, but I would not call it linear. In a true linear there is a working side and a protect side (1+1).
There are a lot of concepts that tend to get muddled together when it comes to the different topologies, just keep chewing away.