Industrial Internet Redundancy: Beyond Simple Backup
Industrial Internet Redundancy is an architectural approach designed to eliminate single points of failure (SPOF) in an industrial site’s WAN connectivity. While a simple 4G backup provides a second path in the event of an outage, redundancy goes a step further:
- Two active links simultaneously (not just in standby mode)
- Transparent failover without interrupting TCP sessions
- Continuous SLA monitoring to measure and verify availability
- Load balancing to optimize the use of both links
- Integration into the BCP (Business Continuity Plan) documented
The goal is to achieve a network availability of 99.99% (i.e., less than 52 minutes of downtime per year), compared to 99.9% (8.7 hours per year) for a single link.
Industrial Network Redundancy Topologies
The Eziwan router supports several WAN redundancy topologies, from the simplest to the most advanced:
Topology 1 — Active/Passive Dual WAN (failover): one primary link (fiber) and one backup link (4G) in hot standby. Automatic failover within <8s. Estimated SLA: 99.99%. Cost: low (1 router, 1 4G backup link).
Topology 2 — Active/Active Dual WAN (load balancing): two simultaneously active links, with traffic distributed between them. Estimated SLA: 99.99%. Bonus: doubled bandwidth. Cost: moderate.
Topology 3 — VRRP with Two Routers: Two physical Eziwan routers (master + slave) sharing a virtual IP address. Redundancy at the router level (not just at the link level). Estimated SLA: 99.999%. Cost: high. Recommended for mission-critical sites.
Topology 4 — Triple WAN: Carrier 1 fiber + Carrier 2 fiber + 4G as a fallback. Estimated SLA: 99.9999%. Used by highly critical sites (data centers, energy providers).
VRRP: Uninterrupted Session Continuity
The VRRP (Virtual Router Redundancy Protocol, RFC 5798) is the industry standard for network gateway redundancy. How it works in the Eziwan context:
IP virtuelle : 192.168.1.1
┌─────────────────────────────┐
│ │
[[Master Router] ●──────── Fiber (active) │
192.168.1.2 │ │
│ LAN industriel
[Routeur Esclave]●──────── 4G LTE (standby) │
192.168.1.3 │ │
└─────────────────────────────┘
The PLCs, HMIs, and LAN devices point to 192.168.1.1 as their gateway—they do not know whether they are communicating with the master or the slave. If the master fails, the slave takes over the virtual IP address in less than 3 seconds. TCP sessions (VPN, SCADA) survive the failover without interruption.
WAN Load Balancing: Bandwidth Optimization
In active/active mode, the Eziwan router distributes traffic between the two links according to configurable rules:
| Strategy | Principle | Recommended Use |
|---|---|---|
| Round-robin | Alternates session by session | Homogeneous traffic |
| Weighted | Proportional to each link’s throughput | Asymmetric links (100M fiber + 20M 4G) |
| By protocol | SCADA over fiber, VPN over 4G | QoS priority by application |
| By IP source | PLCs over fiber, IT over 4G | OT/IT segregation |
The weighted strategy is the most common in the industry: if fiber offers 100 Mbps and 4G offers 20 Mbps, the router sends five times as much traffic over fiber as over 4G.
SLA Monitoring and Reporting
Continuous network SLA monitoring is an essential component of industrial internet redundancy. The Eziwan portal monitors the following in real time for each WAN link:
Availability: uptime/downtime by hour, day, week, and month. Comparison with your provider’s contractual SLA.
Latency: Average, minimum, and maximum RTT to reference measurement points. Detection of latent degradation (slow connection before it drops).
Jitter: variation in latency. Critical for real-time applications (industrial VoIP, cyclic protocols).
Packet loss: the percentage of lost packets. A rate greater than 0.1% indicates congestion or link instability.
Failover History: Detailed log of each failover event, including the timestamp, affected link, duration, and cause.
This data feeds into your PCA reports and helps justify contractual penalties with service providers.
Integration into the Industrial Network Business Continuity Plan (BCP)
Network business continuity is an essential component of the Business Continuity Plan (BCP) for industrial sites. Eziwan’s Internet redundancy enables organizations to meet the most stringent BCP requirements:
RTO (Recovery Time Objective): maximum recovery time. With Eziwan, the network RTO is 8 seconds in Cold Standby mode and 3 seconds in VRRP mode—well below the 1-minute threshold required by most industrial business continuity plans.
RPO (Recovery Point Objective): the maximum acceptable data loss. With local buffering of SCADA and IoT data in the Eziwan router, the RPO is zero: no data is lost during an outage (the data is retransmitted when the connection is restored).
Documentation: Eziwan’s SLA reports and event history serve as documentary evidence of BCP compliance for your internal and external audits (ISO 22301, IEC 62443).