Filter
Top Products
57
Appendix J: Server-to-Switch Distributed Trunking
Distributed Trunking is a link aggregation technique where two or more links across two switches are
aggregated together to form a trunk. This feature overcomes the limitation in IEEE 802.3ad that
specifies all links of a trunk have to be from a single switch. Distributed Trunking improves resiliency
and load balancing in a layer 2 network.
Distributed Trunking (DT) is included in switch software starting with version K.14. In this initial
release, only Server-to-Switch Distributed Trunking is supported. For each downstream server, it sees
the aggregated links as coming from a single switch, which makes any servers that support standard
802.3ad interoperate with Distributed Trunking.
Distributed trunks can be grouped together by configuring two individual DT-LACP trunks with the
common trunk group name. DT ports will be aggregated dynamically after the configuration. The
server/switch should support standard 802.3ad LACP on the links connecting DT switches. It is
assumed LACP on the servers are configured manually on the server with a list of links to be part of
the LACP trunk.
DT offers load balancing of traffic over multiple physical links. From the server to the switch, the
traffic is balanced according to the load-balancing scheme configured on the server NIC. From the
switch to the server, traffic is balanced according to the MAC DA and SA pair.
Two DT switches can be connected via only one Inter-Switch Connection (ISC) and DT ports can be
connected only to servers. Following is the topology supported.
Limitations/Restrictions
• Meshing and Distributed Trunking features are mutually exclusive.
• Routing and Distributed Trunking features are mutually exclusive.
• IGMP and DHCP snooping, ARP-protect, and STP are not supported on DT trunks.
• Q-in-Q in mixed VLAN mode and DT are mutually exclusive.
• ISC ports will be part of all VLANs—i.e., they will become members of a VLAN once that VLAN
is configured.