Network command

I thought that you have to specify the full class network address of the subnet you want to add in EIGRP. So for 10.1.2.0/24 it becomes 10.0.0.0 and for 172.23.6.7 it becoves 172.23.0.0 and voor 192.168.4.5 is becomes

192.168.4.0...

Not sure, though.

schreef in bericht news: snipped-for-privacy@74g2000cwt.googlegroups.com...

10.2.3.0 is correct

10.0.0.0 is a diffrent network altogether

CCNA Nerd wrote:

10.2.3.0 is ok. EIGRP is classless. Take the smallest possible in order to have advertisements going out on only the interfaces you want. You can do network 10.0.0.0. however, you may not want to have neighborship on an interface that has IP address 10.2.6.0 configured, for example. And, make sure the network statement does 'cover' the subnets you configure on the interfaces.

If, however, you want to advertise a loopback address of 10.2.3.5/32 only. Then of course you will need "network 10.2.3.5 0.0.0.0".

1. Use the smaller range when possible
2. Make sure the ranges match for connected peers on connected subnets

Good point in another message - EIGRP is classless. On that note, use this version of the command for effect: router eigrp 17 network 10.2.3.0 0.0.0.7 That way you will set the absolute range of just 0 - 7 in that last octet. The other provided answer was to use the whole class C range: router eigrp 17 network 10.2.3.0 0.0.0.255 That wastes adresses, but let's move on.

If your three routers are connected via Ethernet (10/100/whatever) to each other, they all are connected on the same subnet (same VLAN). They all have to have the same network statement for the range, you cannot have one router missing the network statement or else they will not peer. Don't get into the mindset that if one router advertises that range which they all connect in that the other ones do not have to.

TIP: Want to see the config starting from your EIGRP section? Try this: show running-config | begin router eigrp

Enjoy! Study! Succeed!

1 test away from CCNP!

snipped-for-privacy@gmail.com wrote: [snip]

It's mostly meaningless for EIGRP. Ideally, try to be more specific so you won't run EIGRP on interfaces you don't want to run EIGRP. Remember, unlike BGP, eigrp/ospf both use NETWORK statement to tell the router what interfaces should run the routing protocol.

Yup. I tend to do one of two things with network statements in OSPF. Either very specific or very broad.

If the network has been designed and addresses properly I go for broad. E.G.

net 10.0.0.0 0.0.255.255 a 0 net 10.1.0.0 0.0.255.255 a 1 net 10.2.0.0 0.0.255.255 a 2 net 10.3.0.0 0.0.255.255 a 3

Which will then put any interface that begins 10.0 into area 0, 10.1 into area

1 etc.

If the network has not been designed properly I go very specigic;

int fas 0/1 ip add 10.1.1.1 255.255.255.128

router ospf 10 net 10.1.1.1 0.0.0.0 a 0

and so on.

I don't get why people make it difficult and go

int fas 0/1 ip add 10.1.1.1 255.255.255.128

router ospf 10 net 10.1.1.0 0.0.0.127 a 0

Paul Matthews wrote: [snip]

Just to "over engineer" it for the sake of being fancy!

I find that with OSPF, at the ABRs, you have to be careful if you run multiple areas. But in EIGRP, I don't know that it matters to much.

Couldn't agree more about being specific or very broad.

Given the people who I see do it, I suspect there is a degree of ignorance, and they think that is what they need to do by inverting the mask. One "expert" within an organisation can do quite a bit of damage...

I thought that you have to specify the full class network address of the

10.0.0.0 and

biggest

difference

The wildcard mask option was added in IOS 12.0(4)T for the EIGRP network statement.

Just to "over engineer" it for the sake of being fancy!

I find that with OSPF, at the ABRs, you have to be careful if you run multiple areas. But in EIGRP, I don't know that it matters to much.

Couldn't agree more about being specific or very broad.

Swings and roundabouts TBH. I like all RPs being consistent with each other and networks added under the RP section of the config rather than having to lookat various bits to see what OSPF is doing - I'd rather the interface type bit be under the main OSPF config.

On the other hand, it's kinda nice seeing all the ospf stuff in one place. But I didn't know this was even possible.

[snip]

I do like the above though. When did it become available? Not being able to do show commands while in config mode was always a pain in the ass.