Overview

Configuring QinQ (802.1ad) double VLAN tagging on a non-native ONU requires careful attention to encapsulation mode, TPID settings, and uplink port trust boundaries. This FAQ delivers direct, engineer-grade answers for B2B network architects integrating non-native ONUs into carrier Ethernet, enterprise core, and datacenter leaf-spine environments.

Frequently Asked Questions

Q1: How do I configure QinQ double VLAN tagging on a non-native ONU?

You configure QinQ on a non-native ONU by enabling 802.1ad encapsulation globally, setting the outer TPID to 0x88a8, and applying a QinQ service instance on the subscriber-facing port. Then map the inner C-VLAN to an outer S-VLAN using a VLAN translation or QinQ profile, and set the uplink port to trust the outer tag while stripping or preserving inner tags as required. Verify the configuration with show qinq or display dot1q tunnel, depending on the CLI vendor syntax.

Q2: Which optical modules are compatible with a non-native ONU for QinQ uplinks?

Non-native ONUs typically support standard MSA-compliant SFP, SFP+, and SFP28 modules, including 1000BASE-T, 1000BASE-SX/LX, 10GBASE-SR/LR, and 25GBASE-SR. Third-party modules are generally compatible if they are unclocked or if the ONU allows coding byte override, but you must match the wavelength, reach, and DOM support to the ONU firmware release. For carrier-grade QinQ, always validate the module against the ONU's optical budget and TPID transparency requirements before deployment.

Q3: What is the maximum throughput and backplane capacity supported for QinQ traffic on a non-native ONU?

The maximum QinQ forwarding throughput on a non-native ONU is typically line-rate at 1 Gbps, 10 Gbps, or 25 Gbps per uplink, with a backplane capacity ranging from 20 Gbps to 100 Gbps depending on the model. Because QinQ adds a 4-byte outer tag, the effective payload per frame drops slightly, but the ONU ASIC still forwards at wire speed if jumbo frames and hardware offload are enabled. Check the datasheet for the exact switching capacity and Mpps rating before designing aggregation ratios.

Q4: How does QinQ on a non-native ONU differ from previous-generation VLAN stacking?

QinQ on a non-native ONU differs from legacy VLAN stacking by using standards-based 802.1ad outer TPID 0x88a8 and a dedicated S-VLAN space, rather than proprietary or single-tag 802.1Q encapsulation. This separation allows up to 4094 S-VLANs each carrying 4094 C-VLANs, delivering 16 million service instances, while legacy stacking often limited the outer tag to the same 802.1Q TPID 0x8100. The result is cleaner carrier handoff, better interoperability, and no collision between customer and provider VLAN IDs.

Q5: How do I configure stacking interfaces on a non-native ONU with QinQ enabled?

To configure stacking interfaces on a non-native ONU with QinQ enabled, first assign the physical uplink ports to a stacking port group, then enable QinQ globally and apply the same S-VLAN profile across all stack members. You must ensure that the stacking links carry the outer TPID transparently and that the master election does not disrupt QinQ forwarding tables. After the virtual chassis forms, verify that QinQ service instances are synchronized across all member ports using the stack configuration sync command.

Q6: How does redundant power failover affect QinQ forwarding on a non-native ONU?

Redundant power failover on a non-native ONU should not affect QinQ forwarding if the device has dual hot-swappable AC or DC PSUs and the ASIC retains its forwarding state during the transition. In practice, a failover event causes less than 50 ms of interruption if the ONU supports hitless power redundancy and the QinQ tables are stored in non-volatile memory. If QinQ traffic drops during failover, check PSU alarm logs, verify both power feeds are active, and confirm the ONU firmware supports stateful failover for 802.1ad tunnels.

Q7: How do I troubleshoot QinQ double tagging when the outer VLAN is not being added?

If the outer VLAN is not being added, first verify that the ingress port is configured as a QinQ customer port with the correct S-VLAN mapping and that the global 802.1ad TPID is set to 0x88a8. Common causes include the port being left in access mode, a missing QinQ profile binding, or the uplink port stripping the outer tag before transmission. Use packet capture on the uplink to confirm the double-tagged frame, then check show running-config interface for the exact QinQ service instance and TPID values.

Q8: What QoS and buffer considerations apply to QinQ traffic on a non-native ONU?

QinQ traffic on a non-native ONU should be mapped to QoS queues based on the outer PCP bits, with strict priority or weighted round-robin scheduling for time-sensitive services. Because double tagging increases frame size by 4 bytes, you must enable jumbo frames and verify that the packet buffer can absorb bursts without dropping S-VLAN-tagged frames. For carrier Ethernet compliance, configure per-queue shaping and ensure that the ONU honors the SLA per outer VLAN rather than per physical port.