Re: Portfast in Trunk port

Joe Astorino @ 07/05/2012 12:44 -0300 dixit:
> The main application is just to avoid the normal STP process to allow
> access to the network faster (No 30 second delay on plugging in).
> Generally, it is only used when the device connected is something that
> definitely will not cause a loop -- VMware hosts perhaps that you
> trunk to, IP phones, other such widgets that talk 802.1q but are
> actually end host devices, etc ... if you are doing this towards a
> switch you better make damn sure there are no physical loops.

...or that you do not filter BPDUs there.
That's why portfast continues to send BPDUs even if normally those are
pointless. It's just a safety mech to disable postfast in the not so
happy event that you do connect, e.g., two portfast ports.

-Carlos

>
> HTH
>
> On Sun, May 6, 2012 at 7:31 AM, Carlos G Mendioroz<tron_at_huapi.ba.ar> wrote:
>> Your customer should not be bridging those interfaces, so one going down
>> will not "change the world" for STP.
>> If he is doing a LAG, he should take care of migrating the MACs to the
>> surviving link, usually by imposting some traffic.
>>
>> What kind of "STP inconsistent information" are you envisioning?
>>
>> -Carlos
>>
>> CCIE KID @ 05/05/2012 02:39 -0300 dixit:
>>
>>> Hey Marko
>>>
>>> Yes My SP has two customer facing interfaces and the customer facing
>>> interfaces are trunk ports and they configure Portfast on these
>>> interfaces.
>>> Port fast configured interfaces still receive BPDUs but tat port doesnt
>>> participate in STP Transition states. So my Customer Access laer goes
>>> down,
>>> this info will not be propogated in STP and there may be inconsistent
>>> information in MAC Address table formed by STP
>>>
>>> Can u explain me exactly what is the actual application ?
>>>
>>> Thanks guys for ur thoughts :)
>>>
>>> On Sat, May 5, 2012 at 5:22 AM, Marko
>>> Milivojevic<markom_at_ipexpert.com>wrote:
>>>
>>>> On Fri, May 4, 2012 at 11:00 AM, Anthony Sequeira
>>>> <terry.francona_at_gmail.com> wrote:
>>>>>
>>>>> I am pretty sure this could be beneficial in a router on a stick
>>>>
>>>> scenario. So that of course begs the
>>>>>
>>>>> question - who the ^%&$ would do that in production.
>>>>
>>>>
>>>> If you're an SP - customer-facing interfaces in the access layer, for
>>>> example :-)
>>>>
>>>> --
>>>> Marko Milivojevic - CCIE #18427 (SP R&S)
>>>> Senior CCIE Instructor - IPexpert
>>>>
>>>
>>>
>>>
>>
>> --
>> Carlos G Mendioroz<tron_at_huapi.ba.ar> LW7 EQI Argentina
>>
>>
>> Blogs and organic groups at http://www.ccie.net
>>
>> _______________________________________________________________________
>> Subscription information may be found at:
>> http://www.groupstudy.com/list/CCIELab.html
>>
>>
>>
>>
>>
>>
>>
>
>
>

-- 
Carlos G Mendioroz  <tron_at_huapi.ba.ar>  LW7 EQI  Argentina
Blogs and organic groups at http://www.ccie.net