Help to resolve broadcast storm

[c-deboni]

Good morning,

I am a computer scientist, I changed several months ago all the computer equipment of my company. That’s 20 Catalyst C1000 switches.

For two days now I see that my heart of the network 3 Switch Catalyst C1000 in H-stack falls at 8H interval.

I set up "storm control" on all the copper ports of my switches as well as on the network cores but neanmoin it falls and I am forced to restart it manually.

I did a wireshark capture today at and it crashed about 14:55.

Someone will be able to direct me I can not help myself, I think the problem is a broadcast storm.

[M02@rt37]

Hello @c-deboni 

Do you enable PortFast on edge ports (connected to end devices) to prevent STP recalculations when end devices connect/disconnect ? Also BPDU gurad on these ports too ?

Also, ensure STP is correctly configured and running on all your switches.

Share the storm-control configuration please. If the current storm control settings are not sufficient, consider making them more aggressive. Lower the threshold levels or change the action to shutdown or trap.

[MHM Cisco World]

sorry I check pcap there is no broadcast 
which traffic you see in pcap unusual

MHM 

[c-deboni]

Enable PortFast on edge ports (connected to end devices)
to prevent STP recalculations when end devices connect/disconnect?
Also BPDU gurad on these ports too?

No I don’t have PortFast enabled on any copper port for the final periods.

Also make sure that STP is properly configured and works on all your switches.

I cannot set up STP because my architecture is star and I do not have a second link on the switches located in bay.

Please share the storm control configuration. If the current storm control settings are not sufficient,
consider making them more aggressive. Lower threshold levels or change the action to stop or trap.

Storm Control -> Shutdown
Broadcast Threshhold Level ->50
Multicast Threshhold Level ->50
Unicast Threshhold Level ->50

On all Copper port but nothing on fibre port

@ Cisco World MHM

If it is not a broadcast tempete, I notice neanmoin that a regular interface a few hours one of the switches of the HSTACK freezes (green lights that are fixed and do not flash anymore).

I have nothing on the H STACK logs

Do you have any test ideas I can run?

[MHM Cisco World]

hello Friend 
sorry I was off last two days 
NOW 
the STP must prevent any L2 Loop, 
and for storm since it happened each 8 hours this periodic behave is mostly from Win server send wake up or ask IP of all win clients. 
I will re analysis the pcap file and see which server IP send this broadcast 

MHM