[prev in list] [next in list] [prev in thread] [next in thread] 

List:       ms-ospf
Subject:    Re: [OSPF]
From:       Acee Lindem <acee () redback ! com>
Date:       2008-07-24 14:50:27
Message-ID: D7975380-5550-4CB0-9073-251A318AB816 () redback ! com
[Download RAW message or body]

Hi Guoqiang,

On Jul 24, 2008, at 9:38 AM, guoqiang 03022 wrote:

> Hi, Acee
>
> Thanks for your answer, but i still have a question about the  
> flushing algorithm.
>
> As you say, in most cases, a type-7 prefix is imported by one NSSA  
> ASBR, so there is only one ABR will translate the prefix to Type-5  
> LSA, even though there area more than one ABRs have the translation  
> capability.
> If so, what's the purpose of the flushing optimization?
>
> As per RFC3101, 3.3 Flushing Translated Type-7 LSAs:
>
>  --                                                              
> Conversely
>  --  Type-5 LSAs generated by translating Type-7 LSAs are not  
> immediately
>  --  flushed, but are allowed to remain in the OSPF routing domain  
> as if
>  --  the originator is still an elected translator.  This minimizes  
> the
>  --  flushing and flooding impact on the transit topology of an  
> NSSA that
>  --  changes its translators frequently.
>
> If the two ABRs are going to translate the same Type-7 LSA, then  
> one must flush its translated Type-5 LSA, or
> the other will not do the translation. So the Type-5 LSAs area  
> still flushed and generated if the NSSA changes its trnslators  
> freuently.
> Is there something I misunderstand?

If you've delayed purging your type-5 LSAs until a new translator has  
taken over and originated functionally equivalent translated type 5  
LSAs, then you've achieved your intended purpose. However, I can see  
a race condition here where the new translator has a lower router ID  
than the previous one. This probably should have been mentioned in  
RFC 3101.

Thanks,
Acee


>
> Thanks and Regards
>  guoqiang
>
> _______________________________________________
> OSPF mailing list
> OSPF@ietf.org
> https://www.ietf.org/mailman/listinfo/ospf

_______________________________________________
OSPF mailing list
OSPF@ietf.org
https://www.ietf.org/mailman/listinfo/ospf
[prev in list] [next in list] [prev in thread] [next in thread]