Basic concepts of rapid spanning tree
1, BPDU is the packet that used for communication between switches
2, Root bridge is the switch which has lowest value of priority, mac address
3, Switch port can be root port, designated port, alternative port and backup port; in Switch ports can be categorized also as edge-port(port-fast in classic spanning tree) and point-to-point port.
4, ports statues: discarding, learning, forwarding
5, hello interval is by default 2 secs, max age is 3 * hello interval, thus 6 sec by default
6, when TCN received,switch will immediately timeout its CAM (mac address table), and put all of its none-edge ports in sync mode.
7, During sync, each port will negotiate its new statue with the peer, switch who has the superior BPDU will put port into designated port.
8, Every bridge generate and send BPDU every hello time interval, but only designated port will send out BPDU, not alternative and backup port.
9, Not like classic stp,in rstp bridge will not relay BDPU from root bridge, instead, it will generate its own BPDU with its own BID and Root ID from root bridge. Thus all bridge will know RootID.
10, The sync process is used to determine if I m agree with you about the root bridge. sync process start with sending/receiving proposal BPDU and end by receiving/sending agreement BPDU
Topology change in rapid spanning tree
Scenario 1:Topology change on edge port (up or down)
It will not generate topology change in the network
Scenario 2: Topology chanage on NONROOT switch, linkup on designated port in switch A
1, At first both switches’s ports are in blocking mode until they received each other’s BPDU. If switch A received a inferior bpdu, it will send out BPDU proposal, after got BPDU agreement from peer side, switch A will put port into forwarding state.
2, If switch A received a inferior bpdu, it will send out BPDU proposal, after got BPDU agreement from peer side, switch A will put port into forwarding state.
3, If switch A received a superior BPDU, it will realized that peer switch should be root bride, A will start RPST convergence:
a) Peer switch will send BPDU proposal for negotiation, switch A will block all its none-edge ports, put them into sync mode, age-out CAM (MAC address table) immediately, then reply peer switch BPDU agreement. After that, switch A and peer are forwarding the traffic;
b)All of the none-edge ports in switch A will send out proposal BPDU towards their peer side, and the peer side switch will repeat the step a) for port negotiation.
Scenario3: Topology change on NONROOT switch, linkdown on designated port in switch
1, Swich will stop receiving BPDU from peer side, after 3 x hello interval, switch will think link is down;
2, If the BPDU the switch stopped receiving is inferior BPDU, switch has no need to do anything;
3, If the BPDU the switch stopped receiving is superior BPDU, switch will take that root bridge is lost, if switch has a backup port, that port will immediately be forwarding, it switch has not a backup port, it will start RSTP convergence process. It starts as a wave of handshakes rippling outwards towards the periphery of the network.