[c-nsp] unicast storm

Phil Mayers p.mayers at imperial.ac.uk
Thu Apr 19 03:26:28 EDT 2012


On 04/19/2012 05:00 AM, ujjwal maghaiya wrote:
>
> Could anyone tell to me the possible cases of UNICAST STORM. 		 	   		

One common cause is a host that receives a lot of traffic, but doesn't 
send it - e.g. a syslog server.

If the ARP timeout is > FDB timeout, when the FDB timeout expires, the 
packets will be flooded as unknown-unicast.

Either:

  1. Cause the host to emit traffic
  2. Lower the ARP time to < FDB timeout

Similar things can occur in HSRP setups on the standby route/switch.

If you can be a bit more specific about the symptoms you're seeing, 
people can probably make better suggestions.


More information about the cisco-nsp mailing list