XGS1930 LLDP & CPU High

Hi @SmileitsMe,

I am happy to hear your network is stable. Since the root cause of the disconnection is because of the network loop in your network, you have solved the loop event, right?

The LLDP & LLDP Med made your switch CPU high because your network's topology changed frequently. When LLDP MED topology-change-notification and LLDP notification are enabled on switches, and a topology change occurs, the system logs LLDP notifications at a rate of 7 times in a second. This significantly impacts CPU resources, causing CPU usage to spike to 100% for several seconds.
But this is not the root cause that causes your network to disconnect. The reason I asked to turn off the LLDP notification is that the event log was fulfilled with "LLDP Remote topologyChange" which blurred our focus.

You can feel free to enable LLDP & LLDP Med for sure, since the root cause of the disconnection is not related to LLDP & LLDP Med but the network loop. In normal scenario, transmit/receive LLDP won't cause network unstable.

Hi @SmileitsMe,

You cannot see which port has a loop is that the event log was fulfilled with the "LLDP Remote topologyChange" message. This message washed away the loop detect logs.

I found the loop issue from the previous tech support > show loopguard blank. You may reference my previous reply. I posted the screenshot in it.

Please notice that the root cause of the VM disconnection is not because of the LLDP or CPU usage high, but because of the network loop. Since you have enabled the loop guard, the switch will block the port that has a network loop. Although you have set the errdisable recovery for the loop guard, the network loop still persists before you solve it.

Hi @SmileitsMe,

Thanks for the update. Could I know the detailed information about the network performance wobbling?

By the way, please collect two tech support, one is LLDP & LLPD MED disabled and one is LLDP & LLPD MED enabled for 2 - 3 minutes before collecting, so we can compare the difference.