User guide

Configuring Network Services Protection Switching-EPSR
518
AlliedView NMS Administration Guide
6.13 Protection Switching-EPSR
6.13.1 Overview of EPSR Topology
In Ethernet-based layer 2 Metropolitan Area Networks (MAN), Spanning Tree Protocol (STP) is normally used to provide
redundancy to achieve high availability and continuous access to resources. The iMAP and Allied Telesis Guides explain in
detail how STP works and how it is configured.
Starting in AlliedView NMS release 4.1, the GUI can be used to configure another protection switching scheme, the Ethernet
Protection Switched Ring (EPSR). EPSR provides a 50 milliseconds switching time for an Ethernet-based ring network, similar
to that provided by the Synchronous Optical Network (SONET) protocol. This allows traffic to be redirected around a
faulty link in a ring network fast enough to result in an uninterrupted multicast service (such as video).
As the name implies, EPSR protects only those parts of the network that have a ring topology. Each node on the ring will
have two Ethernet ports connected to the ring. EPSR operates over these Ethernet ports. Key components that are
configured are Control VLANs, Domains, and Protected VLANs.
A Control VLAN is configured on the set of devices, and is used to send and receive control messages over the ring network.
The devices that are included in the control VLAN make up the Domain of the control VLAN.
The VLANs that require fault protection are configured on all the ring ports and are assigned to the EPSR domain. These
VLANs are called Protected VLANs.
Note: There is only one Control VLAN per EPSR domain and it must use tagged frames. This Control VLAN is unique to this domain
and cannot be re-used for another domain.
Note: Control messages use the iMAP Automatic Protection Switching (TAPS) protocol. TAPS protocol control messages are
transported around the ring network for an EPSR domain via its control vlan. This is handled internally by the AlliedView NMS.
The protection scheme basically operates by having an EPSR domain on the ring. The vlans that require fault protection are
configured on all the ring ports and are assigned to the EPSR domain. The control ring determines if there is a loop, in which
case it blocks traffic on the protected VLANs to prevent the loop. If there is no loop, it allows data traffic to flow in either
direction.
6.13.1.1 Master and Transit Nodes
One of the nodes in the ring is designated as the Master node while all the other nodes are designated as Transit nodes. One
ring port on the master node is designated to be the Primary Port (PP) and the other ring port is designated to be the
Secondary Port (SP).
When the ring is operating normally, the master node blocks its SP port for all non-control traffic (data carried over the
protected vlan[s]) belonging to the EPSR domain, preventing a loop on the ring. The layer 2 Ethernet switching and learning
mechanisms operate normally on each of the nodes in the ring. However, the control vlan traffic is not blocked at the SP
port and is allowed to flow through, because the control messages originate either at a master node or transit node but
always terminate at the master node.
When the master node detects a physical link break in the ring, it unblocks its SP port and allows the flow of non-control
traffic through the EPSR domain. Once the master node determines that the break in the ring has been restored, it goes back
to its normal operating procedure.
6.13.1.2 Example Ring Topologies
A typical topology has all devices (or certain ports on those devices) included in the protection domain. Moreover, each
device belongs to only the one domain. However, more complex topologies are possible, as shown in the following figure.