why use backbone fast?

rakem

I have been doing some playing around with different STP modes in my lab.

I'm having a bit of trouble understanding why you would use backbone fast on switches that support Rapid spanning-tree. I don't think you would really.

With normal spanning tree it took 48 seconds for a blocking port to go into forwarding when an indirect link failure occurred.

When i enabled backbone fast on the all the switches that time came down to 32 seconds.

But then rapid spanning-tree was enabled on all the switches the change was pretty much instant. The blocking port when straight into forwarding when an indirect link failure happened.

So im guessing that if your switches support Rapid spanning tree then backbone fast is pretty useless.

is that accurate?

Netwurk

Backbone fast isn't useless, it's built-in to RSTP. So there's no need to enable it manually.

According to Cisco, "You do not need to configure backbone fast with RSTP because the mechanism is natively included and automatically enabled in RSTP"

More info here:

http://www.cisco.com/warp/public/473/18.html

rakem

yea just saw that while watching the Train Signal BCMSN vid. It says that backbone fast, uplinkfast and portfast are all built into RSTP


Cheers.

Netwurk

Portfast can be turned on or off on a per-port basis. The other two are built-in to RSTP.

APA

The uplinkfast, backbonefast features were integrated into 802.1d STP after seeing how effective they were in 802.1w RSTP....... Althought RSTP is the approved method sometimes your switches may not support RSTP so hence making STP support it was the next logical step.....

UPlinkfast does away with the forward delay.....

Backbonefast trims default STP convergence of 50 secs to 30 secs by elminating the max-age-timer(20secs) allows redundant link to go straight to listening mode(15secs) then Learning mode (15 secs) this keeps the forward delay....