Loop Prevention?

Hi everyone I was trying to find out the term for the mixing of different communication packets on a common line like ethernet is called.... when I came across this article on how packets have a max hop count:
"Sometimes packets can get lost and keep bouncing around from router to router, never quite getting to their destination.
A system had to be developed to deal with this because eventually the network would choke with these 'lost' packets. So to solve this problem a 'hop' count is also added to the packet header. Each packet is allowed to 'hop' from one router to another a maximum of say 100 times. Each time the packet passes through a router the ‘hop number’ is decreased by one. If the packet hasn’t arrived at its destination within the number of ‘hops’ allowed then it is deleted by the next router."

And so I'm curious, from my understanding, these days companies don't use Frame Relay because it's so cheap to use dsl or cable, right? Well anyway, can anyone give me an explanation as to how long the TTL is on lost packets and how they are finally dealt with on what I assume to be a "layer 2" communication over the internet?
It seems in CCNA our conversation on WAN ends at the service provider and so I'm left wondering how the internet keeps itself so clean from rogue packets and such.

Referenced site: Teach ICT What is Packet Switching?

Find more posts tagged with

Free for TechExams community: Cybersecurity salary guide

Compare cert salaries and plan your next career move

Button

Comments

networker050184

The TTL in the IP packet is dealt with at layer 3 (it is a field in the IP header) not layer 2 which is why layer 2 loop prevention like STP is so important.

When A packet is routed at each hop the TTL is decremented by one. Once the TTL reaches 0 the packet is dropped on the next hop. The max TTL is 255 and I believe the default for most applications is recommended to 64. So, if a packet is looped it will eventually be dropped whether that is on the internet or inside your corporate network.

You can see this by setting up a simple routing loop in your lab (static route an address at each other on separate sides of a link) then ping the address and you will see the TTL expired message.

danyzuko

I tried dropping a couple of routers and using ip route 0.0.0.0 0.0.0.0 each exiting the serial to the other router and the ping won't loop. it just returns once as successful.

networker050184

Probably because you are pinging an address with a directly connected route?

danyzuko

I put a 3rd router and did a static to each successive router, 3 different subnets, then ran a ping to the 3rd router ip.
The ping spun around 3 or 4 times then just hung there for about 30 seconds and stopped. I was hoping to see an expired message but it was obviously a loop.
I'm scared to see what happens if I do that in GNS3.

EDIT: Actually it didn't work but I found this site that illustrated it well:
https://www.youtube.com/watch?v=lTvCFEZJDhI I guess you have to use a distance vector protocol and turn off split horizon to get it to loop

Thanks for he explanation of TTL anyway.

networker050184

No you don't have to use any routing protocol. This is completely separate from routing information sources. This is strictly forwarding related.

Why are you scared to do it in GNS3?

danyzuko

Even without routing protocols, the ping is successful even with 3 routers with ip statics going round robin.
I was just referring to GNS3 freaking out and freezing my computer but I tried it there too and no dice.

networker050184

If you are doing it in PT it might not happen. It is a simulator after all. GNS3 should work fine, and it won't do any harm. The TTL is there to make sure of that!