Skip to main content

ETHERNET RING PROTECTION SWITCHING WITH MILEGATE

THE MİLEGATE ERPS (ETHERNET RING PROTECTION SWITCHING) SOLUTION PROVIDES:

  • Sub-50ms channel switching for Ethernet ring network Topologies
  • System-level ERPS functionality for Ethernet Switch units of MileGate family chassis
  • Support for master/sub-ring configurations for different inter-ring bonding scenarios
  • Multiple ERP states for multiple logic ring configurations
  • Compliance with ITU-T G.8032v2

Continuous Communication Systems (MCS) need to work extremely well in existing communication networks. KEYMILE is capitalizing on emerging opportunities in communication technologies and influencing the way MCS networks solve their communication needs in packet-switched transmission networks. KEYMILE’s MileGate ERPS solution addresses the evolution of MCS networks from the current TDM to packet-switched Ethernet networks and the challenges MCS network operators face in providing a similar amount of transmission resistance as in TDM transmission networks (e.g. SDH). ERPS provides reliable switching downtimes within 50ms over Ethernet networks. In addition, the MileGate ERPS solution is extensible to a large number of Ethernet gates with capabilities to support multiple physical/logical inter-ring configurations.

ITU-T G.8032V2 OVERVIEW

ITU-T G.8032v2 (ERPS) specifies Ethernet layer network ring and protection switching meanisms. Ethernet rings offer wide-area multipoint connectivity more economically due to the reduced number of links. The mechanisms and protocol defined in this standard enable highly reliable and stable protection of Ethernet networks. It also prevents loops that can severely affect network operations and service.

In an Ethernet ring, without congestion and with all Ethernet ring nodes idle (i.e., with no detected faults, no active automatic or external commands, and only accepting “NR, RB” R-APS messages), with a fiber ring circumference of less than 1200km and less than 16 Ethernet ring nodes, the switching completion for a fault on a ring should be less than 50ms. Under all other conditions on Ethernet rings, the switching completion time may vary to accommodate concurrent APS requests and for compatibility.

erps-rerouteIn Ethernet ring networks, a ring link is bounded by two neighboring Ethernet ring nodes, and a gate for the ring link is called a “ring gate”. The number of Ethernet ring nodes in an Ethernet ring is at least two. Loops in Ethernet rings are avoided by guaranteeing that traffic flows through only one of the ring links. This special link is called the “ring protection link” (RPL) and under normal conditions this link is closed, i.e. not used for service traffic. An engineered Ethernet ring node, the RPL owner node, is responsible for closing traffic at one end of the RPL. In the event of an Ethernet ring failure, the RPL owner node is responsible for opening the end of the RPL, allowing the RPL to be used for traffic if the RPL has not failed. The other Ethernet ring node adjacent to the RPL, the RPL neighbor node, may take part in closing the RPL end. The occurrence of an Ethernet ring fault results in protection switching of traffic. The Ring Automatic Protection Switching (R-APS) protocol is used to coordinate protection actions on the ring.

ERPS considers recyclable and non-recyclable operations. In a recyclable operation, the traffic channel is returned to the running transmission state (RPL closed) after the condition(s) that caused the switching have been removed. In the case of clearing a fault, the traffic channel is returned at the end of the WTR counter. The WTR counter is used to avoid switching protection states when errors occur at intervals. In irreversible operations the traffic channel continues to use the RPL after a switching state has cleared, if no error has occurred.

KEYMILE’s ERPS solutions support recyclable and non-recyclable mode operations and different configurations are possible for each ERP instance defined on the Ethernet ring.

SUPPORT OF MULTIPLE ERP INSTANCES BY ERP

An ERP instance is responsible for maintaining a subset of VLANs. VLANs forward traffic over the physical Ethernet ring. Each ERP instance is independent of other ERP instances that can be defined on physical Ethernet rings.

KEYMILE’s ERPS solution supports Multiple ERP instances (up to 12 ERP instances for each MileGate Chassis Switch). This makes it possible to support ERPS “per VLAN set”, load sharing and interconnected rings.

ERPS INTEGRATED RINGS

Ethernet rings support multiple ring/ladder networks consisting of Ethernet rings articulated with one or more connected points according to the following principles:

  • Main Ring: Consists of a closed ring controlled by its own ERP instance with its own RPL.
  • Sub-Ring: A ring that is connected to a parent ring or a higher sub-ring and does not contain a closed ring. Controlled by its own ERP instance with its own RPL

For larger networks, MileGate ERPS ring connectivity features (including Master Rings and Sub-Rings) are used to extend services across a network. For example, as illustrated, Master Rings and Sub-Rings can be interconnected to support more diverse networks.

Note: Care should be taken when placing EPRS ring links with shared nodes to avoid traffic loops.

In addition, EPRS can leverage any other technology to provide resilient single node or dual-homed connectivity to an aggregation network

ERPS’ SUPPORT FOR POINT-TO-POINT SERVICES

In general, ERPS can support point-to-point (P2P) services. This is done by simply creating a logical ring of two nodes and closing the channel at both ends of the RPL – at the same time making it possible to utilize network resources to the fullest extent.

Furthermore, a logical ring of two (2) nodes can be created on physical networks with multiple nodes (>2).

As can be seen from the figure, a network can contain six nodes {A, B, C, D, E, F}; but only nodes {A, B} can be defined to be the ERPS logical ring. Intermediate nodes {C, D, E, F} They are not involved in ERP and are open to R-APS and CCM messages between nodes. These intermediate nodes can be bridges/switches with point-to-point VLAN cross-connections or even with a separate layer (e.g. SDH).

ERPS’ MULTI-POINT SERVICES SUPPORT

ERPS supports multipoint services as illustrated.

For a rooted multipoint service, leaf nodes can only communicate with designed root nodes, while a designed root can communicate with other leaf nodes.

In order to provide the asymmetric communication needed by nodes in rooted multipoint services, it is possible to assign root-to-leaf traffic in a separate ERP instance from leaf-to-root traffic.

KEYMILE ERPS SOLUTIONS

MİLEGATE CHASSIS SWITCH ETHERNET UNITS

KEYMILE’s ERPS solution can be applied to all MileGate Chassis Switch Ethernet units. That is, any of the Ethernet ports that are part of the chassis switch can be authorized as an ERPS port and defined with many ERP instances (up to 12 ERP instances for each MileGate Chassis Switch) for the configuration of the ERP’s interconnected rings (Master Rings and/or Sub-Rings).erps-subrings

MileGate Chassis Switch Ethernet Units can detect SF status based on loss of physical connectivity (LOS) or selectively based on loss of continuity (LoC) and through Y.1731 OAM CCM messages. The CCM mechanism can detect signal level degradation and link failures in the presence of fi-ber media repeaters/converters, which may occur in some networks.

KEYMILE’s ERPS solution, MileGate Chassis Switch Ethernet Units: Ethernet-over-SDH functionality (EoS) provided by NUSA1/NUSA1-F is qualified and complemented by the ability to forward Ethernet traffic over SDH networks.

SUMMARY

KEYMILE’s ERPS solution is helping MCS evolve and expand into packet-switched transmission Ethernet networks. In addition, MileGate’s ability to transmit ERPS through SDH networks offers complete freedom in the choice of transmission interfaces.

KEYMILE is available on all MileGate Chassis Switch Ethernet units and is a highly flexible and optimized ERPS solution. This allows to build many physical/logical ring network configurations with high availability performance (e.g. sub-50ms switch exchange and recovery, similar to SDH protection mechanisms and much better than other spanning tree protocols) supporting point-to-point and multipoint services.