I understand this principle, but there is no logic with the scenario
On convergence as outlined in BSCN as follows....
R2
R5 ---- R4 ---- R3 ------------R1
Here, R2 has serial links to R1 and R3. Let's say R3=DR and R1=BDR.
BSCN states that if the Ethernet link is disconnected between
R1 and R3, then R1 would sense that the DR went "down" and proceed
To promoting itself as the new DR.
I must admit I haven't tested this exact scenario yet, but I've seen
Similar cases where the two routers in question would get stuck in
Exstart/Exchange forever, and this particular behavior I've succeeded
In reproducing on a test lab.
Elmer
-----Original Message-----
From: Peter Van Oene [mailto:[EMAIL PROTECTED]]
Sent: Monday, September 17, 2001 9:42 AM
To: [EMAIL PROTECTED]
Subject: Re: Need clarification on BDR promotion to DR [7:20163]
So far as the multiaccess segment is concerned, when you remove the DR from
the segment, the DR is down. The only way a DR can do its job is if it has
access to the segment in question. In this case, promoting the BDR makes
sense since the BDR is the only other router on the segment which has
adjacencies with all other routers and has sufficient information on where
the DR left off to be able to take over quickly. When the old DR comes
back, it will not become the BDR unless it is the only other router eligible
on the segment since a BDR election would have already taken place to
replace the promoted BDR.
Does that help?
Pete
*********** REPLY SEPARATOR ***********
On 9/17/2001 at 8:55 AM Elmer Deloso wrote:
>Hi, all.
>I'm trying to understand the PURPOSE and LOGIC behind OSPF BDR
>Promotion to DR. Let's say R1 is DR and R2 is BDR connected via Ethernet
>Link. If I disconnect the cable, this would mean the BDR will promote
>itself
>To DR status, even though the DR never went "down". So when I reconnect the
>Link the DR will just "abdicate" the role? I just don't see the logic here.
>So now the DR will become the BDR, and if we repeat the process of
>disconnecting
>The link, the original scenario is restored. I've read OSPF on CCO and
>Doyle
>but
>Have not come across the explanation of why OSPF was designed to behave
>this
>Way. Any enlightenment on this is welcome.
>
>Elmer
Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=20176&t=20163
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
