EIGRP SIA, Query Behavior

[subramaniyakarthic]

Question1

---------------

I recently read about the behavior of how EIGRP handles queries.

http://www.cisco.com/en/US/tech/tk365/technologies_white_paper09186a0080094cb7.shtml#queryrange

Here,as per the table given when Router one recieves a query from both two and three,which are successor and any FS

neighbor (not the current successor)	passive	reply with current successor information
successor	passive	attempt to find new successor; if successful, reply with new information; if not successful, mark destination unreachable and query all neighbors except the previous successor

R1------R2-----R4----R5---192.168.3.0

     \                   /

       \              /

          R3---/

R2 is sucessor R3 is feasible successor

If R3 queries R1 for the route 192.168.3.0/24 according to the table since R1 has the route as passive and has a route via R2,

will R1 reply to the query from R3 that "I have a route via R2"

Also I have a general question

If  3.0 nw is gone what will be the state of the topology table in R1,R2,R3,R4 before R5 queries??

Will R5 update that the network has gone down  using an update packet to R4

                                  (or)

Does R5 queries all neighbors before even sending an update that a network has gone down??

If this is the case then my first question is valid.

"When R1 recieves a query from R3 which is not a successor,it replies that it still has the route and replies to R3 which in turn replies to R4 and then R5"

Please help with these concepts!!

Or if possible give an explanation of the query concept in a complex scenario

[Peter Paluch]

Hello,

If R3 queries R1 for the route 192.168.3.0/24 according to the table since R1 has the route as passive and has a route via R2,
will R1 reply to the query from R3 that "I have a route via R2"

Yes. Essentially, R1 does not say "I have a route via R2" because that would suggest revealing a part of the topology which is not possible with distance vector protocols like EIGRP. The R1 merely responds with its current distance from the 192.168.3.0/24 which is obviously computed as a metric of the route going through R2, R4 and R5. R3 will add the metric of the link between R3 and R1 to the metric received in R1's reply and it knows now what is the metric to the destination via R1.

If  3.0 nw is gone what will be the state of the topology table in R1,R2,R3,R4 before R5 queries??
Will R5 update that the network has gone down  using an update packet to R4
                                  (or)
Does R5 queries all neighbors before even sending an update that a network has gone down??

R5 would send an Update only if the metric of the network 192.168.3.0/24 changed in such a way that R5 would consider this network stil reachable. However, if the network goes down, R5 does not have any feasible successors that would provide the next shortest path to the disconnected network. Therefore, it has to send a Query packet to all its neighbors to ask for a replacement route. This query will contain the increased metric of the network after R5 noticed that the old route is not reachable anymore. In the case of network going down, the metric is set to infinity. So in your example of network 192.168.3.0/24 going down, only the Query is sent, no Update.

"When R1 recieves a query from R3 which is not a successor,it replies 
that it still has the route and replies to R3 which in turn replies to 
R4 and then R5"

I apologize, I am missing the context of this last question. Can you rephrase it?

Best regards,

Peter