Networking: Ring Topology Protocols
REP or DLR
One ring to rule them all… Many of us are familiar with the famous J.R.R Tolkien quote from his classic The Lord of the Rings. But can this concept also but true in the world of Industrial networking? As always, the answer is: it depends! Let’s consider a few popular ring topology protocols – REP (Resilient Ethernet Protocol) and DLR (Device Level Ring). This post will help explain those ring topology protocols, how they work, and some of the reasons why a facility might choose one ring or the other.
Before we inspect the rings closer, there are a few key terms we need to understand.
- Resiliency – the ability for a network to recover from a fault (break in the cable, missed connection, etc.)
- Convergence Time – the time it takes for a network to “heal” itself
- Requested Packet Interval (RPI) – how quickly data is updated over a connection
It is also important to note that resiliency is NOT the same as redundancy. In a truly redundant network, you would have multiple physical paths and equipment resulting in ZERO convergence time. For most applications, this type of topology is not necessary. However, if your application calls for this type of redundancy (medical, aerospace, etc.), I would investigate Parallel Redundancy Protocol (PRP) and using the 1756-EN2TP ControlLogix® module.
Not sure exactly what that means or exactly where to start? This six-page PDF provides an overview of the standard Allen-Bradley® communications modules, including those that support CIP™ networks (Ethernet/IP, DeviceNet™ and ControlNet™) as well as other network protocols.
Now let's get to the rings.
Resilient Ethernet Protocol
Pictured: an example of a possible REP type ring topology (© Rockwell Automation®)
REP is a Cisco proprietary protocol that provides a way to control loops, handle failures, and help increase convergence time (typically 15ms). This ring protocol is used primarily in Layer 2 CISCO Switch (including Stratix 5700/5400/8000) topologies and does not work on the device level layer. In other words, REP is used between switches only. Additionally, multiple REP rings can exist on a switch. The REP ring is configured by assigning ports certain roles on the switch; Primary, Edge, No-neighbor, No-neighbor Primary, Transit, or none. More information can be found in the reference links below.
Device Level Ring
Pictured: an example of a possible DLR type ring topology (© Rockwell Automation)
DLR is a ring protocol used by modern Rockwell Automation devices such as PowerFlex Drives, Ethernet/IP communication adapters, ControlLogix, and CompactLogix® Controllers, and Stratix® switches. DLR allows automation devices to be placed in a ring with a convergence time of less than 3ms! In other words, if your RPI is above 3ms, it will seem like nothing ever happened on the network – you will only receive a “ring fault”, but not lose connection to your devices. DLR is incredibly simple to set up: you only need to designate a ring supervisor (an actual checkbox in Studio5000) and connect the ring. The ring supervisor simply “watches” the ring and checks for faults.
If there is a case where a device is not DLR compatible (third-party device, older Rockwell device with one Ethernet port), a 1783-ETAP can be added to allow that device to be placed on the DLR ring.
So, which ring protocol should I use?
If you are looking at connecting switches together, REP is the best approach to take. However, down at the device level, where the fastest convergence time is necessary, DLR is your best bet. Based on your equipment, your facility, and your existing network – it depends. Included in this post are resources that may help for further details and clarification. If you need assistance or want to discuss your specific application, reach out to your local PLC/Networking specialist. Contact us today.
- Resource #1). REP Design Guide, Deploying the Resilient Ethernet Protocol
- Resource #2). DLR, EtherNet/IP Embedded Switch Technology
- Resource #3). The Converged Plantwide Ethernet (CPwE) Design and Implementation Guide
- Resource #4). Stratix 5700, Industrial Managed Ethernet Switch
**NOTE*** For even more resiliency, DLR, REP, and multiple Stratix switches can be used in combination with a new Redundant Gateway feature in the Stratix 5700.
Need help decoding networking acronyms? Contact us today!