Unicast Flooding

[clarkmat]

1. Can anyone explain unicast flooding in simple terms? how this occurs and how to remedy the situation, have increased the MAC aging timer from default 300 seconds to 14400 seconds. Basically when one of our server transfers data to a user network our poor 10 Meg users are saturated, is this asynchronous routing? we are using hsrp to the two core 6500 switches with odd/even vlans on different switches to distribute the load.

2. Also what effect does backbonefast have if it is only set on 1 or 2 switches and not all in the network?

Thanks

[pagiles]

Hi,

Look at this.

http://www.cisco.com/en/US/customer/products/hw/switches/ps700/products_tech_note09186a00800a875c.shtml

Basically it can be down to poor design and occurs when you have an instance of Asymmetric Routing.

Paul

[clarkmat]

Thanks for the reply and the url.

When the TCN occurs we lose MAC addresses for a short period of time! Does this mean the TCN flush CAM entries?

This happens to over 200 MAC addresses and then they are re-learnt by the root bridge.

Any ideas?

[glen.grant]

You need to figure out why you are getting TCN's , in a stable network you should be having minimal TCN's . Also make sure any user ports are set up with portfast turned on .

[axfalk]

PMFJI. This link is no longer working. Is there an alternative place to learn about the flooding?

Thanks.

[glen.grant]

Try this

http://www.cisco.com/en/US/partner/products/hw/switches/ps700/products_tech_note09186a00801d0808.shtml