ip cef question

Hey guys.

I am a little confused on two ip cef adj types. What is the difference beween discard and drop?

I suppose what I don't get is what does itmean by prefix is checked? Wouldn't it already be checked if it's checking the adj table?

Discard adjacency
Packets are discarded.

Drop adjacency
Packets are dropped, but the prefix is checked.

Troubleshooting Incomplete Adjacencies with CEF - Cisco Systems

Find more posts tagged with

Comments

Forsaken_GA

You're quoting the exception handling adjacency types.

The drop adjacency basically says, I don't have an adjacency for this, so I'm going to drop the packets, but I'm going to punt to the route processor as well in order to have a shot at building that adjacency. Once the route processor resolves the punt, a full adjacency should come up, but in the meantime, the drop adjacency there is to essentially protect the route processor - you wouldn't want to be able to ddos it by having it constantly punt packets to the RP.

Forsaken_GA

Actually, I may be wrong about that, the drop adjacency may be the one that triggers for unsupported features. I'll need to check my CEF book when I get home

wastedtime

I am no CEF expert but with the help of my CEF book I may be able to. It is saying that Drop Adjacencies will drop the packet but will obey any ICMP requirements. While a Discard Adjacency will not send any ICMP and just silently drop the packet.