Spanning tree works and discovers loops basically by utilizing Bridge Packet Data Units (BPDU). switches send bpdus out the interfaces and by receiving bpdu and assessing the data inside them, they decide if there is a loop of a lower cost link, etc.
in the networks which there is not STP activated on all devices or you have version mismatched on them, this process of exchanging bpdu fails and therefor there wouldnt be a true understanding of the network topology, and u may have loops.
you MUST activate Spanning Tree on all devices, also you MUST run the same version thorugh out ur Layer 2 net. in your case RSTP.
if you run RSTP and Legacy STP on different devices, you get no correct convergence either.
I know that STP use BPDUs to create the TREE etc... But I thought that to just cut is not necesary that the other switch uses STP... I think stp can cut a loop when it detects the same mac in 2 ports...
I thought that all the ports of the other switch have the same mac...
The purpose of STP is to detect L2 Loop and take preventive action when loop is detected.
in your case, when STP is disabled on the switches then you are at a risk of running loops.
regarding the MAC, we have something called a Burnt in MAC (which is unique for the switch) and each interface also has a MAC. The BPDU's uses the Burnt in MAC for its bridge id but to my knowledge, the CAM table learning would be based on interface MAC address
I have 2 switches, one with RSTP enable and the other no.
I have connected it using 2 cables. I have observed that STP doesn't block any port.
Well, the important fact here is whether the other switch with RSTP disabled did at least pass the BPDUs sent from the first switch. It may be possible for a switch with deactivated RSTP to drop received BPDUs altogether. If BPDUs are dropped, STP can not work properly.
I have show the CAM table and I have observed that the ports of the switch that doesn't has STP enable have a diferent mac in each port.... It's logycally that STP not blocks any port...
In order for STP to work, a switch only needs a single MAC address global for the entire device. While Cisco Catalyst switches assign individual MAC address to each switchport, that is not necessary for STP to work. I would not try to deduce anything by just looking at the CAM table regarding STP.
This it's normal? I thinked that STP works because he detected that the 2 ports connect to the same switch...
Yes, it should. However, the 2 ports are detected by a switch receiving, in this case, its own BPDUs. If the other switch discarded those BPDUs instead of flooding them, STP was unable to detect a loop.
STP only blocks one link when the 2 switches have STP enable???
Yes, only one link is blocked. The other remains forwarding, otherwise, the switches would be cut off totally.
In relation to Peter, just wanted to add that a switch has only one BIA MAC address, so if in your case as your are suggesting that 2 switches are connected via two cables and you are seeing two different MAC addresses on those ports, it is simply not possible...
We are pleased to announce availability of Beta software for 16.6.3. 16.6.3 will be the second rebuild on the 16.6 release train targeted towards Catalyst 9500/9400/9300/3850/3650 switching platforms. We are looking for early feedback from custome...