Dell SmartFabric OS10 User Guide Release 10.5.3

LACP fallback in VLT domain

In a VLT domain, LACP fallback enables rebooting of ToR or server that is connected to VLT nodes through VLT port channel. The other end of the VLT nodes is connected to a DHCP/PXE server, as shown in the following figure:

In the above scenario, LACP fallback works as follows:

1. The ToR/server boots
2. One of the VLT peers takes care of controlling the LACP fallback mode. All events are sent to the controlling VLT peer for deciding the port that should be brought up and then the decision is passed on to peer device.
3. The controlling VLT peer can decide to bring up one of the ports in either the local port channel or in the peer VLT port channel.
4. One of the ports, local, or peer, becomes active based on the decision of the controlling VLT peer.
5. Now the ToR/server has one port up and active. The active port sends packets to the DHCP/PXE server.
6. After receiving response from the DHCP server, the ToR/server boots from the TFTP/NFS server.
7. When the ToR/server is fully loaded with the boot image and configurations, the server starts sending LACP PDUs.
8. When the switch receives LACP PDUs from ToR/server, the controlling VLT peer makes the LACP port to come out of the fallback mode and to resume the normal functionality.