STP uplinkfast

[penghon]

Questions about uplinkfast:
For switches with uplinkfast, if the root port goes down, the "standby" link (which is previously down due to spanning tree) goes up within 1-5 seconds. If the root port comes back up, will it do a "preempt" similar to HSRP and cause the "standby" port to go back down again?



Is there any newer stuff to use for layer 2 redundancy other than uplinkfast or spanning tree?

 

[penghon]

And does anyone know what is the difference between uplinkfast and backbonefast?

 

[InDenial]

penghon,

Root port is being selected based on pathcost and port ID. SO when the bad link start working again it will calculate the pathcost to the rootbridge and if there are two ways with the same pathcost to the rootbridge the port with the lowest port ID will be rootport.

Backbonefast:

backbonefast is triggered by het root bridge sending inferior packets wich indicates an indirect link is down.

lets asume three switches. These switches are connected to eachother.

switch x ---- switch y (RB) -/ /- switch z
|(blocking)                           |
---------------------------------------

The link between the Rootbridge (RB) and switch z fails. This means that there is no path for switch z to the rootbridge bacause of the blocked port on switch x. The RB will start sending inferior packets to switch x to say that an indirect link went down.

As a result switch x will put the blocking port into forwarding

so the difference is: uplinkfast is for direct links on a switch and backbone fast is for failing indirect links.

InDenial
CCNA, BCMSN.