Introduction

In modern enterprise campus and carrier-grade networks, the core router serves as the central nexus for inter-VLAN routing and high-speed aggregation. However, improper VLAN trunking configuration on core router platforms often leads to suboptimal forwarding performance, increased CPU utilization, and packet loss at peak traffic intervals (≥ 40 Gbps sustained load). This manual provides a production-ready methodology for configuring IEEE 802.1Q trunking on high-density core routing platforms with backplane capacities exceeding 2 Tbps and sub-microsecond latency targets (≤ 5 μs). We will reference measurable metrics including packet forwarding rate (Mpps), port density (up to 48x100GE), and MTBF (Mean Time Between Failures) thresholds exceeding 300,000 hours.

Core Architecture & Switching Logic

Internal ASIC Pipeline for VLAN Tag Processing

Modern core routers leverage distributed ASIC architectures rather than CPU‑based bridging. For how to configure VLAN trunking on core router effectively, you must understand tag insertion/stripping behavior. Each line card integrates a packet forwarding engine (PFE) with dedicated Ternary Content-Addressable Memory (TCAM) for VLAN lookup. Typical values: TCAM capacity of 128K–512K entries, enabling wire-speed processing of double-tagged (Q-in-Q) frames per IEEE 802.1ad. The forwarding latency for trunk ports remains ≤ 3 μs for 64‑byte frames at line rate.

Backplane Bandwidth and Queuing Mechanisms

A non‑blocking fabric is mandatory for trunking aggregated VLANs. For example, a core router with a switching capacity of 2.56 Tbps can sustain 1440 Mpps (million packets per second) for mixed packet sizes. When configuring trunk ports, ensure that per‑port buffer allocation (typically 2‑4 GB shared buffer) is tuned to accommodate microbursts from multiple VLANs. Industry benchmarks (MEF 3.0) require < 100 ms queuing delay under 95% load; violation indicates improper trunk bandwidth partitioning.

Configuration Workflow: Step-by-Step Production Manual

Step 1: Physical Interface and Sub-Interface Declaration

For a core routing platform (e.g., Nokia SR OS, Juniper JunOS, or Cisco IOS-XR), the first step involves binding a physical 100GE interface to logical sub-interfaces per VLAN. Use the following logic on a typical CLI:

• Set the parent interface to encapsulation dot1q with a native VLAN ID (default often 1, but security hardening recommends disabling native VLAN).
• Define each sub‑interface with a unique VLAN ID (range 2–4094).
• Assign an IPv4 or IPv6 address to each sub‑interface for inter‑VLAN routing.

Concrete example: On a 100GE trunk port, configure 20 sub‑interfaces (VLANs 100–119). Each sub‑interface supports line‑rate forwarding at 100 Gbps with packet size distribution 64–9216 bytes (jumbo frame support).

Step 2: VLAN Filtering and Security Policies

Apply ingress and egress ACLs on trunk sub‑interfaces to restrict traffic between VLANs. For carrier‑grade deployments, integrate MACsec (IEEE 802.1AE) on the trunk link to encrypt all VLAN traffic in hardware, adding only 2–3 μs of latency. Hardware security associations support up to 4096 concurrent secure channels. Without hardware offload, software‑based encryption reduces forwarding performance by >70%.

Step 3: Quality of Service (QoS) Policies per VLAN

Trunk ports aggregate multiple VLANs with different traffic profiles. Allocate guaranteed bandwidth per VLAN using hierarchical QoS: shape VLAN 100 (voice) to 1 Gbps with strict priority, VLAN 200 (data) to 20 Gbps with weighted round robin. Core router buffers (e.g., 40 MB per 100GE port) must be sized to prevent tail drop; enable Random Early Detection (RED) with thresholds at 70% and 90% occupancy.

Technical Specifications

The following parameters are critical when selecting or verifying a core router for VLAN trunking. Always demand datasheet validation against production traffic mixes.

Deployment Scenarios

Use Case 1: Enterprise Campus Core with 10,000+ Endpoints

A university campus core router aggregates 48x25GE trunk ports from distribution switches, each carrying 20–50 VLANs. Total trunk throughput = 1.2 Tbps. How to configure VLAN trunking on core router here requires per‑VLAN spanning tree (MSTP) interoperability and VRRP for gateway redundancy. Measured failure recovery: < 50 ms with BFD (Bidirectional Forwarding Detection) at 10 ms intervals.

Use Case 2: Datacenter Leaf-Spine with VXLAN over VLAN Trunk

Modern datacenters run VXLAN (NVO3) encapsulated within outer VLAN trunks. The core router must support MP-BGP EVPN signaling and handle 256K MAC addresses and 16K VNIs. Hardware VTEP (VXLAN Tunnel Endpoint) line cards achieve 100 Gbps per tunnel with < 10 μs added latency.

Use Case 3: Service Provider Aggregation (Q-in-Q Trunking)

For carrier Ethernet, configure VLAN trunking with S-VLAN (service VLAN) and C-VLAN (customer VLAN) using IEEE 802.1ad. A core router such as those with 10 Tbps fabric and 48x400GE ports can terminate 4000 S‑VLANs, each with 4094 C‑VLANs. Performance: 100% line rate even with double tags, measured by MEF 48.1 criteria.

Conclusion

Configuring VLAN trunking on a core router demands deep hardware awareness—from ASIC pipeline behavior to buffer architecture and security offload capabilities. The presented manual provides a vendor-agnostic framework applicable to any carrier-grade platform. Always validate your configuration with synthetic traffic generators (e.g., Ixia or Spirent) to confirm line-rate forwarding across all trunked VLANs, and implement telemetry monitoring for real‑time VLAN-level packet drops and latency distribution. For enterprises migrating from legacy Layer 3 switches, core router trunking reduces inter‑VLAN latency by up to 80% (e.g., from 200 μs to 40 μs) and eliminates head‑of‑line blocking through per‑port dedicated queuing.