Ruijie RG-S2900G-E Series, Руководство по настройке

Страница: 301 / 943

Configuration Guide QinQ Configuration
QinQ Configuration
Introduction to QinQ
For QinQ, as specified in IEEE 802.1ad, there are so many names in the
industry, for instance, dot1q-tunneling, Tag in Tag, VLAN VPN and Stack VLAN.
Since the VLAN Tag domain defined in IEEE 802.1Q has only 12 bits for VLAN
ID, the device supports up to 4094 VLANs. In real applicatioin environments, for
example, especially in MAN, a lot number of VLANs are necessary for
separation of users. 4094 VLANs is not enough to address this requirement.
The principle of QinQ is that a packet is encapsulated with the VLAN tag of the
network of an ISP before arriving the network and the original VLAN tag iin the
packet serves as data, so that the packet travels the network with two tags. The
packet is propagated in the ISP
’s network by outer VLAN tag (or the VLAN tag
of ISP
’s network), which is stripped when the packet leaves. Then the packet is
propagated in the private network by the VLAN tag of the private network.
As shown in Figure 1, the packets from Network A
’s VLAN 1001 are added with
the outer VLAN tag 1005 before entering the ISP
’s network. Hence, the packets
carry with two tags and be propagated in the ISP
’s network by the outer VLAN
tag 1005. The outer VLAN tag 1005 will be stripped when the pakcets leave the
ISP
’s network. In Network B, the packets are propagated by VLAN tag 1001.
Figure 1 QinQ sketch map
The following figure illustrates the course of adding two tags. The ingress of
edge device is dot1q-tunnle port (or abbreviated as tunnel port). All frames
entering the edge device are considered to be untagged, no matter wheter are