Shouldn't the BGP session itself never have been formed - with a duplicate
Router ID error being the cause. Error code 2 (OPEN error) subcode 3 (bad
BGP identifier) would be the notification message sent.
I can't confirm this, but it would seem to be the most logical protocol
design choice. Forget the decision process - you won't even get that far.
Regards,
Adrian
""Howard C. Berkowitz"" <[EMAIL PROTECTED]> wrote in message
news:p05001914b60b7c7ebd85@[63.216.127.98]...
> >
> >
> > I have one on the BGP Route Decision Process. In pages 168 &
169 of
> >"Internet Routing Architectures," it gives the steps through attributes
it
> >takes when deciding between multiple routes to a destination. I wont list
all
> >of them but if you have the book you know what I'm talking about. If
every
> >attribute matches all the way down the list to the last one, the router
with
> >the highest IP or Loopback address will be the chosen one.
> >
> > This I understand perfectly and I know that it usually wouldn't
even
> >get this far down the list but I find myself again playing the devil's
> >advocate. What if the loopbacks on both devices were the same? I know you
> >could just change one but let's say you didn't. How would it finally make
> >it's decision? Let's pretend everything was the same...would it just,
ummm,
> >pick one? Would it kinda just spin a wheel and see what it lands on kinda
> >like telco's do to give an RFO? ;)
> >
> >Thanks ahead of time for responses folks...it's appreciated. Love your
show.
> >
> >Mark Zabludovsky ~ CCNA, CCDA, 1/4-NP
>
> In formal testing methodology, as defined in ISO 9646, there are
> three kinds of conditions that can be used to test protocols:
> -- correct behavior, typically at the limits of parameters
> -- incorrect behavior, where the packet is errored
> -- inopportune behavior, where the individual packet is correct but
the
> context is wrong for receiving it
>
> What you are describing is an inopportune packet. BGP doesn't
> consider how to handle such--it really doesn't have the information
> to make a decision. Netsys might very well catch a configuration
> error of this type.
>
> The specific response to receiving such an update really would be
> implementation-dependent, but I suspect that most implementations
> would use the most recently received update.
>
> _________________________________
> FAQ, list archives, and subscription info:
http://www.groupstudy.com/list/cisco.html
> Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
>
_________________________________
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
