I have two buildings connected through the CAT6 cable and both ends have cisco-sg350 (capable of etherchannel). Due to the limitations of the cable both the switches are connected through a unmanaged D-link switch. I configured an etherchannel on both the Cisco switches. But the etherchannel is not working because of the unmanaged switch not capable of building the ether-channel. How do I route the lacpdu packets through the unmanaged switch and make the etherchannel work?
As Ron Trunk pointed out, LACP uses a special multicast OUI on its frames. Your unmanaged switch does not recognize the OUI, only that it is multicast, and it sends the multicast frames to every other switch interface. LACP frames are sent with the special
01:80:c2:00:00:02 multicast MAC address. The IEEE has reserved the
01:80:c2 OUI for link protocols, and frames with this special multicast OUI are not allowed to be forwarded from the interface on which they are received to another interface on 802.1D compliant bridges. Obviously, your unmanaged switch is not IEEE 802.1D compliant.
The special OUI will tell a managed switch that the frames are not to be forwarded to any other interface, so putting in a managed switch will not work to allow the two other switches to form a channel. You can use a managed switch in the middle, and you will need to have each switch on the end form a channel with the managed switch in the middle, not the switch on the other end.
How do I route the lacpdu packets through the unmanaged switch and make the etherchannel work?
Etherchannel, LACP or a static LAG trunk can't work across an unmanaged switch ever.
As both Rons have suggested you need either:
- a managed switch with an Etherchannel or an LACP trunk to each of the other switches
- a managed switch w/o LACP filtering (non-compliant to 802.1D)
- two unmanaged switches or two media converters, one for each LACP link (beware: running a static, protocol-less LAG trunk across those will break badly when one of the partial links fails!)
You probably can't with a single switch. LACP frames are multicast, so the unmanaged switch will flood them out all ports. This will cause several problems:
The sending switch will see its frame received on its other port. The switch will interpret this as a loop and refuse to aggregate the link.
The receiving switch will see the same packet on both ports and assume some sort of crossed connection.
You will need to upgrade that unmanaged switch to a managed one and either create VLANs, or run LACP on it.