In XXXX project there was a Network instability in Quidway S7700 and Quidway S5700 series because the STP calculation was triggered constantly and the TC was causing the problem.
XXXX project was a migration for LAN Switches, Core Switch, routers and firewalls. The costumer already had another switches(abbr. S0) and it's suppose there were only switches in the IDF's
Because we couldn't know in specific which port received a cascaded LAN switch, we implemented Edge-port configuration and in additional a BPDU proteccion on each Huawei LAN switch, with this, we could find all of the S0 switches because when the S0 switches started to send BPDU for STP calculation, the port went DOWN because of the BPDU protection.
After that we suggested to the costumer to remove the non-useless switches and also we changed the ports with S0 switches to non-edge port mode.
We were checking the complete topology to check when the problem was occuring or it was in all LAN switches because we were receiving a lot of ARP request and TC in the Core Switch S7700.
After we checked floor by floor all the devices we found that the costumer had more switches out of the topology. That's because there were areas without enough nodes to provide services for all users.
In a new implementation it 's so probably you are not face with this kind of problems, but when you are replacing old devices, you should take care about the hide problems and hide topolgies.