Re: [magma] EXCLUDE (X, Y) to TO_EX (A)

[Mark Fine <mark.fine at alcatel-lucent.com>]

Conceptually, the Router Actions are sequential: first set (A-X-Y) to  
the old Group Timer, then if the router is the elected querier for the  
network, lower (A-Y) to LMQT. If the router is not the elected querier  
for the network, (A-X-Y) would be set to the old Group Timer. If the  
router is the elected querier for the network, (A-Y) would be set to  
LMQT.

The rationale behind the actions are:

(A-X-Y)=Group Timer

(A-X-Y) are the sources to block that were not previously blocked or  
allowed. The router has to indirectly conclude that, despite the new  
request to block these sources, the existing clients want to allow (A- 
X-Y) -- this potentially prevents blackholing traffic from (A-X-Y) to  
existing clients. Since this state is established indirectly, it  
inherits the existing group timer (not the new group timer set to GMI).

Send Q(G,A-Y)

(A-Y) are the sources to block that were not blocked. If the router is  
the elected querier for the network, it has to ensure that no clients  
are interested in these sources before it starts blocking them, so it  
queries for these source while lowering/setting the timers to LMQT.

On Apr 11, 2008, at 11:30 PM, sambath kumar wrote:

> Hi,
>
>   I have a question on the state transitions for this case.
>
> The following are among the actions
>
> (A-X-Y)=Group Timer
> Send Q (G, A-Y) (According to section 6.4.2 2nd Paragraph
>                           when router queries or receives a query, it
>                           lowers the source times to LMQI*LMQT)
>
> This potentially leads to two different values for source timers for
> some or all of (A-Y). Can someone explain which one of the values
> should be used and the rationale behind that.
>
> Thanks in advance for the help.
>
> Regards,
> Sambath
> _______________________________________________
> magma mailing list
> magma at ietf.org
> https://www.ietf.org/mailman/listinfo/magma

_______________________________________________
magma mailing list
magma at ietf.org
https://www.ietf.org/mailman/listinfo/magma