From: Huizinga, Rene (rhuizinga@upcbroadband.com)
Date: Thu Mar 29 2007 - 20:50:15 ART
Hmmm, don't 100% agree. I think both would be counted correct, but if having
to choose between which one would be 'more correct', what would be better
than taking the exact subnet assigned to the interface, assuming that no
'subnet-overlaps' would already be there on the box ??? If the latter would
be the case, then the 0.0.0.0 would be the way to go...
Both generally and practically I'd personally always advise (best practice)
to go for the exact subnet used on the interface targetted and with that,
specify individual commands for each, in combination with the
passive-interface default...
_____
From: Digital Yemeni [mailto:digital.yemeni@gmail.com]
Sent: Friday, March 30, 2007 1:17 AM
To: Jeff Mullan
Cc: Huizinga, Rene; premkumar somasundaram; ccielab@groupstudy.com
Subject: Re: Network 0.0.0.0
the wildcard mask is to specify what interface to send the updates NOT what
network you're advertising! So i guess the 0.0.0.0 <http://0.0.0.0> is the
correct one because it gives more control on what interface you want to use
to send ur update from! 0.0.0255 is just an easy way to include more that
one interface in a single command!
On 3/30/07, Jeff Mullan < jmullan78@gmail.com <mailto:jmullan78@gmail.com> >
wrote:
Actually, this brings up an interesting point . I have always had a feeling
that if we use some thing like 192.168.1.1 <http://192.168.1.1> 0.0.0.0
<http://0.0.0.0> we might lose some points
vs using 192.168.1.0 <http://192.168.1.0> 0.0.0.255 <http://0.0.0.255> ...
can some one pls shed some light on it
pls.
Thanks,
JM
On 3/29/07, Huizinga, Rene <rhuizinga@upcbroadband.com
<mailto:rhuizinga@upcbroadband.com> > wrote:
>
> Hi Premkumar,
>
>
> The network-command under OSPF simply stated is there to define an
> IP-range
> within which the router is scanning all of it's local interfaces that fall
> within it. For all interfaces it finds within that range, it'll ensure
> it's
> taken up into OSPF, within the area configured at the end of the
> network-command.
>
> One can e.g. define an address with a '0.0.0.0 <http://0.0.0.0> '
wildcard-mask to
> specifically identify a single interface instead of using the real subnet.
> But also, as you can see now, if one would use e.g. 'network 0.0.0.0
<http://0.0.0.0>
> 255.255.255.255 <http://255.255.255.255> area 0', the router would take
all of it's interfaces
> configured with IP and place all of them into OSPF, using area 0 for this
> example here.
>
> Can be easy, but at the same time also backfire practically in case your
> network was setup initially like this and you later add e.g. a small
> management-subnet which you didn't actually want into OSPF.
>
> A sample:
>
> Interface fa0/0 has as ip-address 192.168.1.1 <http://192.168.1.1> ,
subnet-mask 255.255.255.0 <http://255.255.255.0> .
> Now all 4 variants stated below will put it into ospf, area 0:
>
> 1. network 192.168.1.1 <http://192.168.1.1> 0.0.0.0 <http://0.0.0.0>
area 0
> 2. network 192.168.1.0 <http://192.168.1.0> 0.0.0.3 <http://0.0.0.3>
area 0
> 3. network 192.168.1.0 <http://192.168.1.0> 0.0.0.255 <http://0.0.0.255>
area 0
> 4. network 0.0.0.0 <http://0.0.0.0> 255.255.255.255
<http://255.255.255.255> area 0
>
> The best option/best practice would normally be to use option 3. It uses
> the
> real subnet configured for that interface. Also it gives you a better
> overview within the router ospf config on which subnets are included.
> Option 1,2 and 4 will do the same but identifies it specifically (1),
> identifies a small range in which the router will also find this interface
> (2) or identifies the whole IPv4 range, risking that other unwanted
> interfaces are taken into OSPF as well.
>
> When wanting to use an aggregate to include OSPF-interface, use at least
> as
> best-practice the 'passive-interface default' command with it and
> 'activate'
> interfaces individually with the 'no passive-interface' command to
> identify
> which ones you'd want to be active and form an adjacency. That won't solve
> the issue of e.g. having the same network being advertised multiple times
> into OSPF in case of local management-links for example, but would at
> least
> be an additional safe-guard.
>
> In favor of using aggregates is the case where one has a nice and tidy
> IP-addressing scheme.
> eg. One would always use the ' 192.168.0.0/16 <http://192.168.0.0/16> '
ONLY for internal core-links
> within your network which should always be active. In that case one can
> include the 'network 192.168.0.0 <http://192.168.0.0> 0.0.255.255
<http://0.0.255.255> area 0' for all core-link
> carrying routers into your config-baseline !!!
>
>
> Cya
>
> Rene.
>
> -----Original Message-----
> From: nobody@groupstudy.com <mailto:nobody@groupstudy.com>
[mailto:nobody@groupstudy.com <mailto:nobody@groupstudy.com> ] On Behalf Of
> premkumar somasundaram
> Sent: Thursday, March 29, 2007 11:24 PM
> To: ccielab@groupstudy.com <mailto:ccielab@groupstudy.com>
> Subject: Network 0.0.0.0 <http://0.0.0.0>
>
> Team,
> Could any one tell me what is the purpose of the command network
> 0.0.0.0under Eigrp or OSPF??...
>
> Prem
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
<http://www.groupstudy.com/list/CCIELab.html>
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
<http://www.groupstudy.com/list/CCIELab.html>
This archive was generated by hypermail 2.1.4 : Sun Apr 01 2007 - 06:35:53 ART