In 802.1Q, what is the difference between IVL and SVL?

  • 3 December 2013
  • 0 replies

Userlevel 3
Article ID: 4918


Vlan learning settings

Unicast flooding

For devices supporting 802.1Q VLAN functionality, there are two possible modes for the operation of the Source Address Table:
  • IVL (Independent Vlan Learning): VLANs each use their own logical Source Address Table. Transparent source address learning done as a result of ingressed vlan x traffic, is not made available on behalf of ingressed vlan ytraffic for forwarding purposes. This setting would be useful for handling devices (such as mainframes) with NICs which share a common MAC address. [/code]
  • SVL (Shared Vlan Learning): Two or more VLANs are grouped to share common source address information. This setting would be useful for configuration of more complex "asymmetrical cross-vlan" traffic patterns, without forcing the switch to flood the unicast traffic in each direction.
This brings us to the concept of a Filtering DataBase ID (aka FDB ID, aka FID). Though the reality is slightly more complex, a FID can be thought of as a plain, garden-variety Source Address Table. Different FIDs = different SATs = Independent VLAN Learning; and same FIDs = same SATs = Shared VLAN Learning.

Depending upon the product, the VLAN Learning setting will be global for the entire switch (SVL or IVL, potentially selectable), or can be configured in a more granular manner in that some VLANs each have unique FIDs while other VLANs share a FID.

A given device typically has a single Source Address Table, with each entry containing a FID parameter which can, if IVL is being used, make it appear that separate SATs are in use.

See 5397 for configuring IVL/SVL on a Matrix DFE.
See 5498 for related Matrix C1/SecureStack C3/C2/B3/B2/A2 information.

0 replies

Be the first to reply!