Re: OSPF LSA type 3 filtering

From: Marko Milivojevic <markom_at_ipexpert.com>
Date: Fri, 4 Jan 2013 02:24:56 -0800

Exactly. Without T2, there is NP-Complete problem and there is no solution in finite amount of time. At least that's how people smarter than me explained it to me :-)

--
Marko Milivojevic - CCIE #18427 (SP R&S)
Senior CCIE Instructor - IPexpert
:: This message was sent from a mobile device. I apologize for errors and brevity. ::
On Jan 4, 2013, at 2:21, Carlos G Mendioroz <tron_at_huapi.ba.ar> wrote:
> Haha, no to the best of my knowledge.
> 
> NP-complete is a kind of algo complexity.
> 
> -Carlos
> 
> Marko Milivojevic @ 04/01/2013 07:16 -0300 dixit:
>> Btw. I believe the problem I was trying to describe is called NP-Complete, but that's a bit outside of my area of expertise. I'm not a mathematician not a CS.
>> 
>> --
>> Marko Milivojevic - CCIE #18427 (SP R&S)
>> Senior CCIE Instructor - IPexpert
>> 
>> :: This message was sent from a mobile device. I apologize for errors and brevity. ::
>> 
>> On Jan 4, 2013, at 2:06, Marko Milivojevic <markom_at_ipexpert.com> wrote:
>> 
>>> What is missing is the decisive information on when you can stop computing. Without T2, there is no knowing when you are done :-). What if an area had 5000 routers.
>>> 
>>> You are correct though. You *could* do it without Type 2, but it can lead to dangerous issues.
>>> 
>>> --
>>> Marko Milivojevic - CCIE #18427 (SP R&S)
>>> Senior CCIE Instructor - IPexpert
>>> 
>>> :: This message was sent from a mobile device. I apologize for errors and brevity. ::
>>> 
>>> On Jan 4, 2013, at 2:02, Carlos G Mendioroz <tron_at_huapi.ba.ar> wrote:
>>> 
>>>> Marko,
>>>> say we have an area with 3 routers, R1, R2, R3, connected by a LAN.
>>>> Then OSPF would choose one as DR. Say that lan is X.
>>>> 
>>>> Would you agree that the database representation would be:
>>>> 
>>>> Router links:
>>>> R1: R1 -> DR (transit)
>>>> R2: R2 -> DR (transit)
>>>> R3: R3 -> DR (transit)
>>>> 
>>>> Net link:
>>>> DR: X (R1,R2,R3)
>>>> 
>>>> You can draw the topology just by looking at the router links.
>>>> What is missing ?
>>>> 
>>>> -Carlos
>>>> 
>>>> 
>>>> Marko Milivojevic @ 04/01/2013 01:11 -0300 dixit:
>>>>> Writing on a phone. Pardon the brevity
>>>>> 
>>>>>> 
>>>>>> I would argue that you can make the topology of an area only with type 1 LSAs, and that type 2 LSAs are just for "condensing" the multiaccess
>>>>>> link reachability information in one place.
>>>>> 
>>>>> Not quite. You would know which routers exist in the area, but not how they are interconnected.
>>>>> 
>>>>> To calculate the SPF tree, routers need two pieces of information for all non-leaf links: the link state, and relationship with other routers.
>>>>> 
>>>>> OSPF recognizes three link types in Type 1: stub, transit, and point to point.
>>>>> 
>>>>> For point to point links, link state is carried in two link state entries. Link itself is described as a "stub link", and the relationship with other router is described as a point-to-point link. These are both in Type 1 LSA.
>>>>> 
>>>>> However, for transit link the actual link is described as a link entry in Type 1 LSA, with a reference to a Type 2 LSA (in a form of a DR address). The Type 2 carries the topological information about the relationships between touters in the segment. Both are crucial for the topological calculation.
>>>>> 
>>>>> Note - this was all about the topological information and not the reachability.
>>>>> 
>>>>> -Marko
>>>>> 
>>>>> 
>>>>> 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
> 
> -- 
> Carlos G Mendioroz  <tron_at_huapi.ba.ar>  LW7 EQI  Argentina
Blogs and organic groups at http://www.ccie.net
Received on Fri Jan 04 2013 - 02:24:56 ART

This archive was generated by hypermail 2.2.0 : Sun Feb 03 2013 - 16:27:17 ART