Which mechanism can be used to avoid a count to infinity loop?

Which mechanism can be used to avoid a count to infinity loop?

a. split horizon
b. route poisoning
c. holddown timers
d. triggered updates
e. split horizon with poison reverse

I pick B. route poisoning, because it marks the route hop count as 16 (unreachable). Am I correct?

Find more posts tagged with

Save $250 on 2025 certification boot camps from Infosec!

Book now with code EOY2025

Button

Comments

NetworkVeteran

Route poisoning--in the absence of hold down timers, triggered updates, poison reverse, or split horizon--is insufficient. Try drawing three routers connected together like triangle. Work through the scenario with each of these options in turn.

workfrom925

NetworkVeteran wrote: »

Route poisoning--in the absence of hold down timers, triggered updates, poison reverse, or split horizon--is insufficient. Try drawing three routers connected together like triangle. Work through the scenario with each of these options in turn.

My current Cisco knowledge doesn't allow me to demonstrate these scenario in detail yet. All I know is, for RIP, route poisoning changes the hop count to 16 when a link fails, without counting the hops one by one, thus avoid the count to infinity. So why wouldn't route poisoning be the corrected answer?

JoshyJ

I would say the answer is a and e.

Route poisoning is to stop routers from routing through bad routes, not to stop loops.

d6bmg

How did you get this question at the first place?

workfrom925

d6bmg wrote: »

How did you get this question at the first place?

Haha. It's from Cisco NetAcademy. Their answer is "holddown timer" with no explanation.

NetworkVeteran

Here's a simple three-router topology:

By now, you are probably immensely awed by my artistic talent and wondering how I did that, but let's focus on distance-vector routing for now. Suppose R1 is connected to 66.6.0.0/16 and has told all the world.

R1: connected
R2: rip/2-hop
R3: rip/2-hop

Suddenly, R1 goes down, and with it that devlish route! Miraculously, R2 and R3 don't see link to R1 going down. (This, of course, implies there are some hidden L1 or L2 devices between the routers.)

R1: --dead--
R2: rip/2-hop via R1
R3: rip/2-hop via R1

Three minutes later, R2 notices R1 is dead! As you may recall, RIP is not known for being a particularly speedy protocol. R2 uses a Cisco-proprietary feature--the hold-down timer. It places the route in the hold-down state.

R1: --dead--
R2: --hold down=179s--
R3: rip/2-hop via R1

Suddenly, R2 receives an update from R3! "Don't fret! I know how to get to 66.6.0.0. It's only 3 hops away via me." If the Cisco hold-down feature were not activated, R2 might accept this route, and then re-advertise it. While "The Count" from Sesame Street would be very happy, network administrators would be very sad. They might even kill a cat!

R1: --dead--
R2: --hold down=179s--
R3: rip/2-hop via R1

Shortly thereafter, even R3 also realizes there's a problem with R1.

R1: --dead--
R2: --hold down=177s--
R3: --hold down=179s--

Epilogue: R1 comes up. Her neighbors don't believe her for almost three minutes!

R1: connected
R2: --hold down=163s--
R3: --hold down=165s--

Note: You'll learn more from labbing this then by reading it.

simple-router-topology.png

NetworkVeteran

You might also find this page on hold down timers interesting--

The TCP/IP Guide - RIP Special Features For Resolving RIP Algorithm Problems

JoshyJ

The question is a weird one. If the question said what mechanism can be used to avoid a count to infinity loop in a redundant network? Then holddown timer being the only answer would hold true.

But if we talk about also a non redundant network then other answers are also correct. Maybe this question is just about which is the MOST correct answer.

NetworkVeteran

Maybe this question is just about which is the MOST correct answer.

"Which answer is most correct?" is a good general way to approach exams.

Don't worry--on the real certification exam, the questions will generally be more cut-and-dried than this. They (and interviewers) will simply confirm that you can explain what each of those feature does and where they might go wrong.

I hope drawing it out helped the OP. Good luck with your studies.