Cisco 15327 - Manual (Page 254)

17-2
Ethernet Card Software Feature and Configuration Guide, R7.2
Chapter 17 Configuring Cisco Proprietary Resilient Packet Ring
Understanding Cisco Proprietary RPR
Understanding Cisco Proprietary RPR
Cisco proprietary RPR is a MAC protocol operating at the Layer 2 level. It is well suited for transporting
Ethernet over a SONET/SDH ring topology and it enables multiple ML-Series cards to become one
functional network segment or shared packet ring (SPR). Cisco proprietary RPR overcomes the
limitations of earlier schemes, such as IEEE 802.1D Spanning Tree Protocol (STP), IEEE 802.1W Rapid
Spanning Tree Protocol (RSTP), and SONET/SDH when used in this role.
In Software Release 7.2 and later, the ML-Series card supports IEEE 802.17b based RPR (RPR-IEEE)
in addition to Cisco proprietary RPR. Throughout this book, Cisco proprietary RPR is referred to as
Cisco proprietary RPR, and IEEE 802.17 based RPR is referred to as RPR-IEEE. This chapter covers
Cisco proprietary RPR.
Chapter 26, “Configuring IEEE 802.17b Resilient Packet Ring” covers IEEE
802.17b based RPR.
Role of SONET/SDH Circuits
The ML-Series cards in an SPR must connect directly or indirectly through point-to-point STS/STM
circuits. The point-to-point STS/STM circuits are configured on the ONS node and are transported over
the ONS node’s SONET/SDH topology with either protected or unprotected circuits.
On circuits unprotected by the SONET/SDH mechanism, Cisco proprietary RPR provides resiliency
without using the capacity of the redundant protection path that a SONET/SDH protected circuit would
require. This frees this capacity for additional traffic. Cisco proprietary RPR also utilizes the bandwidth
of the entire ring and does not block segments like STP or RSTP.
Packet Handling Operations
When an ML-Series card is configured with Cisco proprietary RPR and is made part of an SPR, the
ML-Series card assumes a ring topology. If a packet is not destined for network devices bridged through
the Ethernet ports of a specific ML-Series card, the ML-Series card simply continues to forward this
transit traffic along the SONET/SDH circuit, relying on the circular path of the ring architecture to
guarantee that the packet will eventually arrive at the destination. This eliminates the need to queue and
process the packet flowing through the nondestination ML-Series card. From a Layer 2 or Layer 3
perspective, the entire Cisco proprietary RPR looks like one shared network segment.
An ML-Series card configured with Cisco proprietary RPR has three basic packet-handling operations:
bridge, pass-through, and strip.
Figure 17-1 illustrates these operations. Bridging connects and passes
packets between the Ethernet ports on the ML-Series and the packet-over-SONET/SDH (POS) ports
used for the SONET/SDH circuit circling the ring. Pass-through lets the packets continue through the
ML-Series card and along the ring, and stripping takes the packet off the ring and discards it.
The Cisco proprietary RPR protocol, using the transmitted packet's header information, allows the
interfaces to quickly determine the operation that needs to be applied to the packet. It also uses both the
source and destination addresses of a packet to choose a ring direction. Flow-based load sharing helps
ensure that all packets populated with equal source- and destination-address pairs will be sent in the
same direction, and arrive at their destination in the correct order. Ring direction also enables the use of
spatial reuse to increase overall ring aggregate bandwidth. Unicast packets are destination stripped.
Destination stripping provides the ability to have simultaneous flows of traffic between different parts
of a ring. Traffic can be concurrently transmitted bidirectionally between adjacent nodes. It can also can
span multiple nodes, effectively reusing the same ring bandwidth. Multicast packets are source stripped.