Guidelines and Limitations

This is a list of guidelines and limitations to keep in mind when configuring bridge domains:


  • Currently on bridge domain ports the QoS default behavior is as follows:


    • Incoming 802.1Q frames’ CoS value (a.k.a. PCP) is preserved along with the VLAN tag when transported in the VXLAN payload
    • Incoming 802.1ad frames’ CoS value from the inner tag is preserved along with the inner VLAN tag when transported in the VXLAN payload
    • For outgoing 802.1Q frames after VXLAN decapsulation CoS is set to value 0
    • For outgoing 802.1ad frames after VXLAN decapsulation the CoS value carried in the VXLAN packet is preserved and becomes the CoS value of the inner tag. The outer tag instead carries a CoS value of 0.


  • Currently on the Dell S5232-ON and Dell S5248-ON switch models the QinQ and untagged modes is not supported while the 802.1Q mode  is allowed.


  • On ports added to a BD, the STP protocol and the IGMP snooping features are not supported.


  • Routing on BDs is not supported.


  • vNETs are not supported with BDs


  • The LLDP and LACP protocols are supported on BD ports configured in QinQ mode. The trunking and vLAG functionalities are supported too.


  • The optimized ARP and optimized ND features cannot coexist with BDs. They should be set to off in the system-setting-show output. If the user attempts to enable the feature, a conflict message is printed as below:


CLI (network-admin@switch) > system-settings-modify optimize-arps

system-settings-modify: Cannot turn on optimized ARP as bridge-domain Q-in-Q is configured


When optimized ARP/ND is already configured before a port is added to a BD, a similar conflict message is printed:


CLI (network-admin@switch) > bridge-domain-port-add name bd1000 port 18 outer-vlan 18 inner-vlan 100

bridge-domain-port-add: Cannot add port in Q-in-Q mode to bridge-domain as optimized ARP or ND configured


  • The flowtrace function is not supported with VXLAN and BDs.


  • The VLAN ID is not shown in the l2-table-show output when a L2 entry is associated to a cluster tunnel:


CLI (network-admin@leaf1) > l2-table-show bd bd1 format switch,mac,bd,vlan,vxlan,inrf,ports,tunnel


switch           mac               bd  vlan vxlan  intf ports  tunnel                          

---------------- ----------------- --- ---- ------ ---- ----- ---------------------------------

leaf1            00:12:c0:80:1c:09 bd1 100  101001 274  82                                

ursa-scale-leaf2 00:12:c0:80:1c:09 bd1 100  101001 274  82                            

ursa-scale-leaf1 00:12:c0:80:40:52 bd1      101001            auto-tunnel-10.40.40.1_10.30.40.1

ursa-scale-leaf2 00:12:c0:80:40:52 bd1      101001            auto-tunnel-10.40.40.1_10.30.40.1

ursa-scale-leaf4 00:12:c0:80:40:52 bd1 100  101001 81   81



  • When configuring vLE redundancy, if LACP is used for end-to-end connectivity checks (which is usually preferable), a vLE trunk can be set up only with multiple individual virtual pipes using separate VTEPs (e.g., 8 in case of a 4-way vLE trunk). When using static trunks, instead, it is also possible (but generally less desirable) to set up, say, a 4-way vLE trunk with 2 2-way trunks as endpoints (which equates to a 4 x 2-port trunk configuration on leaf switches and a 4-way trunk configuration on end devices).