CCNA - Question?

Author

Question?

justindu

2003-10-04, 12:04 pm

What methods are used to prevent routing loops in Distance-Vector protocols?

A. Defining maximum number of hops.
B. Split Horizon.
C. Poison-Reverse.
D. Hold-Down Timers.
E. Triggered updates.

ABCE OR ABE

I have seen some sites show poison reverse as an option, and others dont...

Whats the right answer?

Spunknot

2003-10-04, 12:58 pm

I've been studying from the Cisco 640-607 Odom book, and in that book it shows that the answer as seen on the 640-607 would be B&D, however, there is an extra-credit section on that chapter (information that isn't pertinent to the 640-607, which would show the answer as being ABCE. That is why you're finding some differentiating answers on this one.

Hope that helps

edmonds_robert

2003-10-04, 2:40 pm

Poison reverse isn't really used to prevent routing loops. It is used to "poison" a route that is no longer available by advertising it with an infinite metric. For example, in RIP, it would advertise the dead route with a hop count of 16. This would tell the other routers that you can't get there from here.

thomashong

2003-10-04, 2:43 pm

I still think that is the answer...So split horizon and route poisoning should be the answer for this question.

TH

OHCCNP2003

2003-10-04, 3:11 pm

If you have to pick three, go with B, C, D.

Split horizon will not allow a router to advertise a route back out the same interface it got the update from. This will prevent a router (2) from sending the
updated information it received from router (1) back to router (1).

Hold down timer will be tell routers to restrict, for a specific time period, any changes that might affect recently removed
routes. To remove, one of the following must occur:

The holddown timer expires.

Another update is received with a better metric.
A flush timer, which is the time a route would be held before being
removed, removes the route from the routing table.

Route poisoning instructs a router to make a infinite route entry in its routing table. Using rip as an example, this would we a value of 16.

Assuming your taking 640-607, you can set hop count on IGRP I believe out to 255. The default is 100. Rip has a max count of 15 which I don't remember being changeable. I never use Rip or IGRP so I would really have to look it up to make sure.

Triggered updates are not part of distance-vector. They belong with link-state.

Spunknot

2003-10-04, 3:44 pm

quote:
Originally posted by edmonds_robert
Poison reverse isn't really used to prevent routing loops.

This statement is wrong according to the Wendell Odom book. It states:

Issue
Routing loops occur due to updates passing each other over a single link

Solution

Split horizon - The routing protocol advertises routes out an interface only if they were not learned from updates entering that interface.
Split horizon with poison reverse - The routing protocol uses split-horizon rules unless a route fails. In that case, the route is advertised out all interfaces, but with infinite-distance metrics

And there is also Route poisoning, which isn't mentioned here, but that also solves the issues of Routing loops occur due to updates passing each other over alternative paths.

So what do you think now?

edmonds_robert

2003-10-05, 8:12 pm

I think that (1) there's no need to be flippant, this is a discussion forum and (2) I sit corrected.

Spunknot

2003-10-05, 10:14 pm

I think you took my post the wrong way; I wasn't trying to be rude. I just wanted to get things right, and rid us of any ambiguity. I'm sorry if you were offended.

edmonds_robert

2003-10-06, 10:04 am

In that case, I guess I'll let it slide just this once.

Sponsored Links