Here is a nice article describing why in some scenarios (like
redistribution done on broadcast media) network wildcard mask helps
avoiding icmp redirects.
http://www.netmasterclass.net/site/articles/Forwarding%20behavior%20of%20IGP%20routing%20protocols%20on%20the%20broadcast%20part%20I.pdf
-Azher
Roger wrote:
Thanks Scott, I got it. This is the way I was thinking about it.
And sure, my example was wrong, there was a typing error on the second
example...uuups...
greetings
Roger
-----Ursprüngliche Nachricht-----
Von: Tyson Scott [mailto:[EMAIL PROTECTED]
Gesendet: Mittwoch, 14. Mai 2008 23:29
An: L. Jankok
Cc: Roger; [email protected]
Betreff: Re: [OSL | CCIE_RS] Network command
Roger,
It doesn't matter.
But I think it is important to understand what you are really doing.
So just to fix the example
It should be
172.16.1.1 0.0.0.0 area xy
172.16.1.0 0.0.0.255 area xy
Both of these commands are similar but different. The first command
adds specifically only the ip of 172.16.1.1 to OSPF. The second
command adds any interface that falls within the 172.16.1.0/24 network
range into OSPF. So if you have an interface with the network of
172.16.1.0/24, the second command is only going to add that interface
to the OSPF process. (As you can't have overlapping ip's configured on
two interfaces on a router.)
But if you have 172.16.1.0/30, 172.16.1.4/30, 172.16.1.8/30, etc
etc.... the network command will add all the interfaces that fall
within that network range to ospf area xy.
So look at what you are trying to accomplish. If a questions asks you
to put x interface in area x and y interface in area y and they have
similar IP ranges it doesn't matter whether you have the command
applied to the interface or to the process. But if you have a router
in a single area with 20 different interfaces for similar networks it
may save you time to just add all the networks to the ospf process as
a single network range.
This concept is basically the same accross all the routing protocols.
But with protocols like rip that automatically summarize the network
command it then becomes a good test question to not send routing
updates out unecessary interfaces.
On Wed, May 14, 2008 at 4:23 PM, L. Jankok <[EMAIL PROTECTED]> wrote:
It depends what you are asked to do.
I will go for "network 172.16.1.1 0.0.0.0 are xy" for this is more in
line
with the new method of activating the routing process directly under the
interface.. if the requirements are different I will comply.
On Wed, 14 May 2008 22:17:45 +0200
"Roger" <[EMAIL PROTECTED]> wrote:
: Hi Group,
: : I have a general question regarding the mask of the network command
within
: the routing protocol. Let's say you have configured and interface with
the
: address 172.16.1.1/24 and you want to have it in your OSPF. Which mask –
: wildcard do you configure?
: : network 172.16.1.1 0.0.0.0 area xy
: : or
: : network 172.16.1.1 0.0.0.255 area xy
: : Sure, I understand the difference, with the second example I have the
whole
: class c in OSPF, but does it really matter? How is it in the LAB?
:>From my praxis work, I would use always the second example...
: : Thanks
: : Roger
