Hirose: Connecting the future
Industrial Ethernet Book Issue 30 / 31
Request Further Info   Print this Page   Send to a Friend  

Dual-homing: resilience up to the network's edge

While managed Ethernet switches for use in redundant LAN structures such as rings and meshes have become increasingly available and popular, redundancy for edge devices on the network are not quite so well understood and provided for. A new class of dual-homing edge switch fills the void experienced by designers of redundant LANs. This tool can open up LAN design options and expand device choices. It permits any industrial SCADA device to be creatively used as part of a resilient network.
By Frank Madren

Designing and implementing high-availability Ethernet LAN topologies in industrial networks can be challenging. The serial field buses traditionally used for industrial control do not offer fault-tolerant options, and industrial systems designers are not readily accustomed to redundancy planning for control devices. In the past, choices for implementing redundancy for devices at the network's edge were too limited, expensive and complicated to be considered for use in most industrial systems.

While ring and mesh network topologies are coming into general use, finding practical ways to provide for recovery from faults at edge devices and nodes has remained difficult. The software required to manage computers and other devices fitted with dual NICs (Network Interface Cards) to set up redundant connections on the network remains complex, and the dual-NIC solutions are costly. Few manufacturers offer PLCs and Industrial Ethernet devices with dual connectivity built in for industrial sensor and controller applications. We think that dual-homing technology for small industrial Ethernet switches is changing the picture. Redundancy at the edge of industrial networks can be greatly simplified by use of dual-homing switches.

What is dual-homing?

In Ethernet LANs, dual-homing is a network topology that adds reliability by allowing a device to be connected to the network by way of two independent connection points (points of attachment). One access point provides the operating connection while the other implements a standby or back-up connection activated in the event of operating connection failure.

A dual-homing switch, with two attachments into the network, offers two independent media paths and two upstream switch connections. Loss of the Link signal on the operating port connected upstream indicates a fault in that path, and traffic is quickly moved to the standby connection so accomplishing fault recovery.

For industrial applications, the most common redundant Ethernet LAN topology is the ring. Although a ring structure will recover from a media break or the failure of a switch, the failure of a switch in the ring takes down the nodes connected into that switch. Uptime for a critical node can be increased by connecting it to two switches in the ring. Dual-ported nodes or a dual-homing switch can be used to achieve this. The critical nodes so connected to the ring will remain attached to the ring even when one of the two ring switches goes out of service.

Dual-homing improves reliability in two ways when used in rings. First, it uses two media connections into the ring so that media faults cause less down-time. Second, the MTBF of specialised dual-homing units - inescapably a single point-of-failure for the attached nodes - is much greater than that of a typical managed Ethernet switch in the ring.

Dual-homing also provides connection flexibility during network maintenance. A switch in the ring may need to be temporarily taken out of service to load new software or to perform a hardware upgrade. During a planned outage, dual-homing maintains the connection of critical nodes in a redundant LAN to ensure device up-time while permitting service work to be done during normal work time.

Some LAN designs use two rings for greater redundancy. Single Ethernet rings can recover from only one fault at a time. With two rings, there are two paths for LAN traffic so providing the potential for recovery from more than one fault at a given time. Dual-ring configurations may sometimes be achieved at little cost increase by breaking up a single large ring into two smaller rings.

In a dual-ring topology, uptime requirements typically make cost considerations secondary. There are a few special-purpose (and expensive) dual-homing PLCs and dual-NIC computers for installation in complex dual-ring LANs. Incorporating the facility within Industrial Ethernet switches permits any Ethernet-enabled edge device to be installed in a dual-homing configuration.

Dual-homing may be used to connect critical nodes into one ring while providing a failover path into a second industrial Ethernet ring

For instance, to use our own ES42D dual-homing switch in a dual-ring LAN requires connection of Port 1 into any switch in one ring and Port 2 into any switch in the other ring. The fault-recovery logic for handling fault conditions in dual-ring LANs takes the same form as for single rings. Loss of link on the operating port causes the dual-homing switch to move all traffic on the upstream connection to the standby port, effectively reversing the order of the standby and operating ports.

Dual-homing in mesh structures

A mesh structure provides an even more robust redundant LAN. Mesh complexity is almost without limit, with multiple interconnects adding to the fault recovery capability even when multiple faults occur. The cost of the Ethernet LAN equipment and cabling used in constructing the mesh rises as the complexity increases necessitating a trade-off calculation between absolute network resilience and infrastructure cost.

Dual-homing switches can create another type of redundancy for critical nodes in a mesh. Meshes are complex structures and require managed Ethernet switches running complex software such as STP or RSTP. The mesh operation will vary the forwarding and blocking ports for the mesh switches as the network goes through changes and recovery from a fault or faults. Exactly how the switches will operate to recover from faults cannot always be predicted or tested. The reliability advantage of the dual-homing Switch in a mesh is that it is small, simple and predictable. Because it is inherently a much more reliable device than the typical managed mesh switch, it provides a valuable addition for improving the availability of the most critical mesh network components.

Media redundancy

A simple but occasionally useful application for a dual-homing Ethernet switch is to provide for media redundancy. Where a significant risk of media faults exists - for example where the media is attached to a movable device and is flexed during normal operation, or where undependable wireless interconnects are used, or where part of the media path is exposed in a high-security situation - media redundancy adds essential connectivity uptime. For media redundancy, both ports 1 and 2 (the Operating Port and the Standby Port) are connected to the same switch upstream. Media types for ports 1 and 2 may be either copper or fibre, and each port's media type can be configured independently of the other. Additional media types such as wireless may be inserted so long as the Link signal is passed through to provide a reliable signal for switchovers between the switch operating and standby ports.

Frank Madren is president, GarrettCom, Inc.

Source: Industrial Ethernet Book Issue 30 / 31
Request Further Info    Print this Page    Send to a Friend  


Analog Devices: Time Sensitive Networking
DINSpace fiber optic and Cat 6 patch panels
ICP DAS at SecuTech Thailand
Japan IT Week Autumn

Get Social with us:

© 2010-2018 Published by IEB Media GbR · Last Update: 12.12.2018 · 31 User online · Privacy Policy · Contact Us