We have a network infrastructure containing 2 Catalyst 3560's and several Catalayst 500 Express. The 3560's are used ase core switches, the 500's are access switches for phones and desktops.
The 2 3560's are connected directly with two interfaces (copper), no etherchannels of special portsettings. In my opinion, the 3560's act as an active/passive cluster. I would like to have an active/active cluster. Looking at the attached images, I think that NIC 3 and NIC 4 of the server are only connected for redundancy in case of failure of a 3560.
The 500's are connected with the two Gbit interfaces to both 3560's. STP shuts down one of these connections. This ends up in all client traffic redirected to one 3560. This makes me think only NIC 1 and 2 from the server are used. Is this correct ? How can I optimize the situation so all network traffic from the clients is divides between both 3560's and so all server NIC's ?