VLAN Tagging

TPID (16 bits): Tag Protocol Identifier set to 8100 (hex).

P (3 bits): IEEE 802.1P Priority level, set between 0 and 7.

C (1 bit): Canonical indicator set to 0.

VID (12 bits): VLAN ID, set between 0 to 4095.

UP (3 bits): User Priority

CFI (1 bit): Canonical Format Indicator

VLAN ID (12 bits): Virtual Local Area Netword Identiifer

TPID P C VID

Destination

Address Source

Address Type/

Length Data CRC

VLAN Tag

UP CFI VLAN ID

Note: The TPID can actually be thought of as an Ethertype designation, identifying the payload as a VLAN. The original MAC frame’s Ethertype field is moved to the inside of the VLAN payload, following the TCI.

The TCI contains the 12-bit VLAN identification, 3-bit priority field, and 1-bit canoni-cal format indicator (CFI). The VLAN ID can have a value between 0 and 4095.

However, values 0, 1, and 4095 are reserved and best avoided. The priority field allows the network administrator to assign a value from 0 to 7 based on the type of traffic. The CFI is always set to 0 for Ethernet traffic.

User Priority Traffic Type

0 Best Effort

1 Background

2 Spare

3 Excellent Effort

4 Controlled Load

5 Video < 100 ms latency and jitter 6 Voice < 10 ms latency and jitter

7 Network Control

Table 9 User Priority VLAN Membership

Ethernet traffic can be assigned VLAN memberships through several means:

By Port: all traffic through a particular switch port is assigned the same VLAN.

• Fast traffic forwarding.

• Easy to maintain for network administrators.

• VLAN membership tied to geographic location.

By MAC address: Each MAC source address is assigned a specific VLAN ID.

• Great flexibility.

• VLAN lookup tables require manual configuration by network administrators.

• MAC address lookup takes more processing time.

By Protocol: VLAN IDs are assigned based on IP address, or protocol used (such as AppleTalk).

• Great flexibility.

• Protocol lookup takes more processing time.

By Authentication: VLAN IDs are assigned based on authentication credentials or the result of IEEE 802.1X authentication results.

• Improved security.

• Ideal for wireless connectivity.

Stacked VLAN Tags

IEEE 802.1ad amends 802.1q by providing a means to stack multiple VLAN tags for traffic management and bridging. This technique of placing one 802.1q tag inside another is often called “Q-in-Q”.

DA: MAC Destination Address SA: MAC Source Address Etype: Ethernet type/length TAG: 8021.Q.VLAN Tag DATA: Customer Data FCS: Frame Check Sequence

MAC DA MAC SA Etype TAG Etype Data FCS

Length C-Customer

Enterprise CPE S-Carrier

Access To Carrier Core

C-MAC C-Payload Data

C-MAC C-Tag C-Payload Data

C-MAC S-Tag C-Tag C-Payload Data

Figure 58 Stacked VLAN Tags

Stacking VLAN tags is an efficient means of differentiating traffic through a net-work backbone, especially when then the user data may itself have VLAN tags.

The outer tag, also known as the service tag or S-tag is distinguished from the customer tag, or C-tag. The TPID of the C-tag is usually 0x8100, as for normal VLAN traffic. The TPID of the S-Tag may have a proprietary value, depending on the implementation by the vendor. Each tag layer has its own priority setting. The priority of the outer tag allows the network provider to achieve the desired quality of service for the bridged traffic.

VLAN and Frame Size

Because the minimum payload size for an Ethernet frame is 46 bytes, the presence of the 4-byte VLAN TPID and TCI pushes the minimum frame size from 64 bytes to 68 bytes. Likewise, the largest, non-Jumbo frame size increases from 1518 to 1522 bytes. Stacked VLAN tags also increase the minimum and maximum frame sizes by 4 bytes per VLAN tag.

When a device receives a VLAN tagged frame that is only 64 bytes, and it must remove the VLAN tag and forward the Ethernet payload, it is left with a frame that is only 60 bytes long. At this point, the device may simply drop the frame. Some systems may add 4 bytes of filler at the end of the payload to create a legal 64-byte frame.

6.1.7 MPLS

Multi Protocol Label Switching architecture was designed to provide a unified data carrying service/simple routing for both circuit-based clients and packet-switching clients providing a datagram service model. Basically, it allows voice, IP, ATM, Frame Relay and Ethernet services all to be carried on the same network. It can be used with many types of framing, including Ethernet.

The Layer 3 label analysis is only just once, when the packet enters the MPLS domain. After that, labels are just inspected to continue packet forwarding.

Layer 2 Header Top Label

...

Data Layer Network Layer

MPLS

1 bitS Label

20 bits ESP

3 bits TTL

8 bits 4 bytes

Figure 59 MPLS Structure

The MPLS header contains a ‘stack’ of one or more labels. A label has four fields as shown in Figure 59:

• 20-bit Label value.

• 3-bit field for CoS priority (EXP, experimental).

• 1-bit bottom of S (Stack) flag. If used, it signifies the current label is the last in the stack.

• 8-bit TTL (time to live) field; The Time to Live label will expire at the conclusion of this number of time-to-live hops.

• The EXP (Experimental) field can be used to distinguish classes of service, or per hop behavior, for differing classes of traffic traveling within the MPLS tunnel (AKA Label Switched Path - LSP). Alternatively, an LSP carrying a single traffic class uses the label to determine the per hop behavior of the class.

6.1.8 Ethernet Standards and Resources