Stacking feature technology combines multiple physical switches into a single logical entity, enabling centralized management through unified configuration, enhanced reliability via cross-device redundancy with automatic failover, and flexible scalability through linear port/bandwidth expansion—all while maintaining synchronized forwarding tables (MAC/IP) and streamlined operations.
Key Benefits of Stacking:
✔ Cost-effective – No need for expensive chassis switches
✔ Highly available – Redundant architecture with fast failover
✔ Scalable – Expand ports/bandwidth by adding members
✔ Simplified management – Unified control plane for the entire stack
Ethernexion launched this stacking functionality in Q2 2025, now supported in the new E-NOS version V3.0.18 for both S5 and S7 series products.
Technical Background
Ethernet switches come in two form factors: box-style and chassis-style.
- Box-style switches offer low cost but lack high availability and uninterrupted service protection, making them unsuitable for critical scenarios.
- Chassis-style switches deliver high performance, high port density, and high availability, making them ideal for mission-critical deployments—but they come with high upfront costs and elevated per-port pricing.
Stacking technology bridges these two approaches by combining their advantages.
By connecting multiple devices through stacking ports, stacking creates a single virtual logical device. Managing this virtual device allows centralized control of all stack members. This hybrid solution retains the cost efficiency of box-style switches while delivering the scalability and high reliability of chassis-style distributed systems.
How to implement
Stacking Implementation Process
- Physical Connection:
- Member switches are interconnected via dedicated stacking ports using either:
- Daisy-chain (linear) topology
- Ring topology (recommended for higher reliability)
- Member switches are interconnected via dedicated stacking ports using either:
- Role Assignment:
- The system automatically elects one device as the Master (manages the entire stack) and others as Slaves (execute control-plane commands).
- Role election is dynamic and adjusts upon topology changes (e.g., Master failure triggers re-election).
- Data Forwarding:
- Traffic flows through logical stacking links, with deterministic forwarding paths based on the topology.
- Ring topologies enable bidirectional traffic for higher bandwidth utilization.
How to configure
The following network example shows a ring stack system
- Devices: Access switches Switch1-Switch5 form a 100G ring topology via uplink optical ports.
- Purpose: Ensures high reliability for access-layer networks through stacking.
- Stacking Roles:
- Master: Switch1 (priority 200)
- Backup Masters: Switch2 (190), Switch3 (180), Switch4 (170), Switch5 (160)
- Stacking Ports: eth-0-23 to eth-0-26 on each switch.
