If industrial networks fail, the results could be disastrous. While data interruptions are worrisome for any business, sudden gaps in connectivity could put industrial organizations at risk of significant productivity loss or introduce substantive safety concerns. And, with pandemic pressures continuing to impact both supply and demand frameworks, even small network interruptions can have big consequences.
The result? For industrial firms to drive sustained success, robust network redundancy is critical. But not all redundancy approaches are created equal. There’s a need for interoperation over proprietary production. That's why there’s an open protocol approach. In practice, this means industrial switches need to be fully customizable, flexible and easily interoperable with switches from multiple suppliers.
Looking to improve your network redundancy, but not sure where to start? Look at common connective pain points, popular network redundancy protocols and approaches to help your team find the best business protocol.
Also read: Where IT and OT converge
The challenge of continuous uptime
It's a question of when, not if, networks will face the challenge of unexpected disruption. The culprit could be cyberattacks, power failures or weather events that knock connections off-line. Without effective layer 2 redundancy, companies face an uphill climb to get systems back up and running and get production back on track.
Redundancy protocols help to solve this problem by proving a network failover framework; if one node fails, protocols automatically reroute traffic to minimize the impact of downtime. But redundancy itself often presents a challenge when it comes to finding and deploying industrial access points, routers and switches at scale.
Here's why: while some providers may promise the benefit of a substantive switch ecosystem to help boost network recovery, the protocols used by these switches are often proprietary. While this may not pose a problem for companies operating in a limited geographic area, business networking expansion quickly becomes problematic as companies are compelled to keep building out proprietary systems, even if current solutions no longer meet their needs.
Another issue is the need to overcome legacy expectations. Historically, industrial applications in SCADA systems or industrial control systems were entirely internal. Many were air-gapped and lacked any connectivity to public networks at large. Even within intranets, access was strictly controlled to limit the risk of unauthorized use. The advent of the Industrial Internet of Things (IIoT), however, has changed the nature of networks. Now, interconnection is a key facet of operations. From sensors that help to pinpoint problems or notify managers of proactive maintenance needs to Internet-facing devices that interface with supplier and logistics portals, there's no facet of operations that remains untouched by the move to always-on, on-demand connections.
The result is a paradox: Companies need protocols to deliver continuous uptime, but they're often stuck in a system that doesn't permit the flexibility they need to keep pace with changing market conditions.
Popular network redundancy protocols
So, what are your options when it comes to industrial network redundancy protocols?
Spanning tree protocol (STP): Spanning tree remains a popular layer 2 protocol. Most deployments leverage the IEEE 802.1D specification across bridges and switches; spanning tree creates a topology of redundant links on your network while avoiding the creation of loops. These redundant links are similar to file backups in your network. If one link fails, the connected backup links activate so users can continue to work. In practice, spanning tree diagrams often look like spiderwebs; multiple devices are connected to a single switch, which is in turn connected to several switches to create a multi-path failover framework.
Multiple spanning tree protocol (MSTP): MSTP expands the usability of STP to include multiple trees across differing network sites. This is especially useful for companies undergoing growth and expanding into new satellite offices or operational facilities. In practice, MSTP makes it possible for administrators to map their preferred number of virtual local area networks (VLANs) to a single MSTP instance. For example, if you have two distinct layer 2 topologies but six VLANs, you only need to create two MST instances and then distribute your VLANs as desired.
Rapid spanning tree protocol (RSTP): RSTP leverages the IEEE 802.1w specification to both increase availability and reduces the risk of redundant loops. By allowing network traffic to be rerouted around failed nodes, RSTP makes it possible to boost both network performance and availability to minimize overall downtime. It also improves the process of preventing redundant loops by blocking redundant paths on a network. This is critical since redundant paths within a network redundancy solution cause more problems than they solve. Consider the issue of broadcast storms. If switches in a loop configuration detect a failed node and begin broadcasting duplicate data packets, these packets inevitability reach the next node in the loop, which then receives and rebroadcasts the information. The result is a continual cycle of broadcast and rebroadcast that can quickly degrade network performance and overwhelm switches.
Ethernet ring protection switching (ERPS): ERPS offers a different approach to network redundancy. Instead of opting for the spiderweb approach of STP, ERPS uses the ITU-T G.8032 standard to create a ring of nodes that is naturally configured to prevent loop issues. Here's how: While nodes are arranged in a ring, one connection is always blocked to prevent the creation of a loop. This means that traffic can flow in both directions around the ring but always stops at the blocked link. If another link in the ring goes down, however, it becomes the blocked link and the previously blocked link is opened, in turn allowing data flow to continue at the same rate with virtually no loss of speed. ERPS rings can also be connected in multiple layers to create larger stacks that offer greater overall performance than STP. Even over hundreds of miles of fiber connections, the protected ring structure of ERPS means that ping won't drop and connections will remain stable.
How to pick the right protocol
So, what's the best choice for your business? Are you better served by STP or one of its variants, or is ERPS the ideal choice? The answer? It depends.
If you already have an MSTP or RSTP deployment in place and need to add new switches, it's likely more cost-effective to keep the same redundancy protocols. While removing all current STP links and rebuilding redundancy from scratch would eventually result in a more reliable network, the amount of time and effort to achieve this goal often makes it prohibitive for industrial enterprises that rely on continual uptime to maximize productivity.
ERPS, meanwhile, is a great choice if you're developing a new network deployment. With fewer complications and faster recovery times, if nodes do go down, ERPS provides the ideal framework to support mission-critical applications but comes with both time and resource commitments.
It's also worth noting that ERPS is an open-source framework used by many large telecommunication providers, which means that it adheres to carrier-grade standards of performance and interoperability. In addition, ERPS permits link aggregation across networks, which makes it possible to create multiple links across networking switches to boost overall redundancy.
Put simply, the use case, budget and current framework typically determine the best fit for redundancy protocols. While ERPS does offer increased performance and reliability, this may be offset by its initial setup costs and time, especially if existing MST and RST deployments are complex and interconnected. If you're looking at building out a new redundancy framework that prioritizes rapid recovery, however, ERPS may be your best choice.