Solved: Re: Eigrp query

rawatmohinder · ‎03-30-2006

Q:-- which three of the following terms are known as reliable packets in EIGRP ??

A.) Hello

B.) ACK

C.) Reply

D.)Query

E.)Update

choose three..

ruwhite · ‎04-03-2006

That's another trick.... :-)

Ack's are nothing but any other packet with an extra TLV containing the sequence number of the packet being acknowledged, and the ack bit set in the packet header. So, an ack _could_ be reliable, if the ack is contained in an update, which does happen from time to time. But, when you ack a packet, you're never acking receiving the ack, but rather you're acking the other information that might happen to be in the packet.

In fact, we use this specific concept at startup to prevent unidirectional neighbor relationships from being formed. When an EIGRP router received a hello from another router, it will put the new neighbor in "pending" state, and send a unicast empty update. The new neighbor will then ack, but the ack must be contained in an update as well, so the ack is "reliable." After this, the new neighbor will be placed in its normal state, and topology tables will be exchanged, etc.

So, it can go both ways, but we normally don't consider ack's to be reliable, in terms of a test, etc.

:-)

Russ

View solution in original post

pkhatri · ‎03-30-2006

Hi,

The three packets that are sent reliably are:

C. Reply

D. Query

E. Update

Hope that helps - pls rate the post if it does.

Paresh

gopal_4476 · ‎03-30-2006

Hi

Reply,Query,Update are reliable packets in EGIRP.

Hello and ACK are not.

Regards,

Gopal

ruwhite · ‎04-02-2006

Trick question! No fair! :-) There are actually four packet types that are reliable in EIGRP, but if this is from a test, I doubt the test has been updated to reflect the fourth type, which is an SIA-Query.

:-)

Russ.W

pkhatri · ‎04-02-2006

Hey Russ,

How about the Request packet ? Was that ever implemented ?

Paresh

ruwhite · ‎04-03-2006

I know we have some test capabilities, but I don't know if any of them actually use the request or probe packets.

:-)

Russ

bhedlund · ‎04-02-2006

What Russ hasn't mentioned here is that he has written a fantastic article on EIGRP in the latest edition of Packet Magazine.

I encrourage everyone to read it:

http://www.cisco.com/web/about/ac123/ac114/ac173/Q1-06/p_15.html

Thanks Russ.

Respectfully,

Brad

rawatmohinder · ‎04-02-2006

Hi all

Thanks for ur reply...

According to me reply,query & update packets are reliable..

Acknowledgement packets are also reliable in some cases.

Thanks

Mahi

ruwhite · ‎04-03-2006

That's another trick.... :-)

Ack's are nothing but any other packet with an extra TLV containing the sequence number of the packet being acknowledged, and the ack bit set in the packet header. So, an ack _could_ be reliable, if the ack is contained in an update, which does happen from time to time. But, when you ack a packet, you're never acking receiving the ack, but rather you're acking the other information that might happen to be in the packet.

In fact, we use this specific concept at startup to prevent unidirectional neighbor relationships from being formed. When an EIGRP router received a hello from another router, it will put the new neighbor in "pending" state, and send a unicast empty update. The new neighbor will then ack, but the ack must be contained in an update as well, so the ack is "reliable." After this, the new neighbor will be placed in its normal state, and topology tables will be exchanged, etc.

So, it can go both ways, but we normally don't consider ack's to be reliable, in terms of a test, etc.

:-)

Russ

rawatmohinder · ‎04-03-2006

Thanks Russ...

You explained the concept very well...

Thanks a lot..

Regards

Mahi