Priority and Backbone-Fast

[ShyLion]

On page 531 (571 in PDF page numbers) it is said what Backbone-fast increases default priority to 49152, but we enabled backbone-fast on all switches previosly on page 524 (564) and so priority should have been increased on all switches as well, not s1 and s2 only.

Can you clarify on this?

[Fuzz]

Backbonefast does not change the STP bridge priorities, Uplinkfast does. You should only use Uplinkfast on access-layer devices that end-users connect to. Uplinkfast changes the bridge priority to a high number to ensure that the switch does not become the root bridge, and increases all port costs by 3000 to prevent downstream switches chosing this switch to access the root bridge over better paths.

Backbone fast works by finding alternate paths to the root bridge when it receives inferior BPDUs. It sends out RLQ messages in an attempt to find the root bridge, which sends an RLQ reply. Switches that don't have Backbonefast enabled won't reply, and so the whole system fails.

Important to remember: Uplinkfast for access layer switches only; Backbonefast on all switches, including the root switch.

[dae26]

Page 531 says "...BackboneFast raise the default priority to 49152...". Does that not mean there was a change in priority for using BackboneFast?

[Fuzz]

Backbonefast does not alter the bridge priority, it must be a typo because it is Uplinkfast that increases the priority.

Uplinkfast is only used on access layer switches, so the bridge priority is increased to make it an undesirable path to the root. Backbonefast allows fast convergence when the link to the root fails. There is no need for it to alter the bridge priority for this to work, and it is safe to use on the root bridge - how could this be possible if the priority was altered?