network outages with 2 HPE 5130

[dgr44700]

I have a core network with 5130 switches. the 2 switches are stacked with IRF configured. other switches are connected to this core network. Since the installation of these 2 new switches our network is slow and suffers from failures (loss of access to the various servers). I have a GE2 / 0/5 port which puts the message GE2 / 0/5 was notified a topology change every 5 seconds. Is this what can disrupt the entire network? How to avoid this problem? should we activate loopback protection on the ports?
Thank you for your help

[Ivan_B]

Hi @dgr44700 !

Yes, flapping of a single port may cause a havoc in your network if MSTP or RSTP is enabled. This is due to the fact how switches react on a TC event - they flush all MAC address table entries immediately (unlike legacy STP) and this may cause temporary short-term so-called "unknown unicast storm". And if these flapping events are frequent you may end up in a situation when your MAC table is never in the stable state. My recommendation - first, set that port as STP edge port ("stp edge-port" command under the interface context) and monitor if this will improve the stability of your network. Edge ports flapping does not cause STP TC events and therefore their status change does not affect the STP topology and MAC tables of STP devices, that's the difference. Also, set all ports that do not face STP-capable devices as edge ports - it will greatly improve stability in your network.

[Ivan_B]

Hi @dgr44700 !

Did you have time to try the solution proposed? Did it resolve the problem?

Thank you in advance!