RE: OSPF wildcard bits

From: Gyo (Gabor.Gyori@xxxxxx)
Date: Sun Jun 30 2002 - 14:09:03 GMT-3


   
Hmm.. No I can not reproduce this behaviour using 12.1.14.
I have made the previous tests using 12.1.15, on loopback (with ospf network po
int-to-point) and physical interface as well. I am sure that it worked the way
I described because it was a real surprise for me.

So I have to withdrawn the second result. If the IP address of the interface is
 covered by one of the network .. area commad, it advertises it with the specif
ied interface subnet mask regardless the wildcard mask in network .. area comma
nd.

Thanks for correction,

Gabor

-----Original Message-----
From: Colin Barber [mailto:Colin.Barber@telewest.co.uk]
Sent: Sunday, June 30, 2002 4:38 PM
To: ccielab@groupstudy.com
Subject: RE: OSPF wildcard bits

Are you saying that using a 0.0.0.0 mask you end up with /32 networks
advertised? This is incorrect, the networks are advertised correctly. Either
you tested with a extremely buggy IOS or you were testing with loopbacks
which, by default, are advertised as host networks.

Maybe I misunderstood you?

Colin.

-----Original Message-----
From: Gyori Gabor [mailto:Gabor.Gyori@lnx.hu]
Sent: 30 June 2002 11:40
To: Ahmed Al-Ghawas; ccielab@groupstudy.com
Subject: RE: OSPF wildcard bits

It confused me too, so I have made some tests.

The results:
 - The /32 mask is sufficient for the interface in order to run OSPF on it
 - The OSPF advertises the given interface with /32 mask instead of the
   original (f. e. /24 interface subnet mask)
   In fact, the interface is advertised with the lower mask of interface
    mask and the network .. area mask.

So I disrecommend to use /32 mask in ospf process expect it is directly
stated to advertise the interface as /32.

Gabor

-----Original Message-----
From: Ahmed Al-Ghawas [mailto:ghawas@batelco.com.bh]
Sent: Saturday, June 29, 2002 9:37 PM
To: ccielab@groupstudy.com
Subject: OSPF wildcard bits

Guys,

I am really getting confused!

>From what I understood from "CCIE Practical Studies" book that when using
the
network command, that you ought to be precise in the extent that you would
add
the interface ip address and not advertise the whole subnet mask!

For example:

R1
S0:10.1.128.1/24
|
|
S0:10.1.128.2/24
R2
S1:10.1.80.1/24
|
|
S1:10.1.80.3/24
R3

R1 config:
router ospf 2001
network 10.1.128.1 0.0.0.0 area x (and not; network 10.1.128.0 0.0.0.255
area
x!!)

R2 config:
router ospf 2001
network 10.1.128.2 0.0.0.0 area x (and not; network 10.1.128.0 0.0.0.255
area
x!!)
network 10.1.80.1 0.0.0.0 area x (and not; network 10.1.80.0 0.0.0.255 area
x!!)

R3 config:
router ospf 2001
network 10.1.80.3 0.0.0.0 area x (and not; network 10.1.80.0 0.0.0.255 area
x!!)

I really need to stick to one concept and understand why this book doest it
differently then the other and avoid loosing marks for such stupid thing in
the real lab!!

Any input is much appreciated

Ahmed



This archive was generated by hypermail 2.1.4 : Tue Jul 02 2002 - 08:12:44 GMT-3