Mailing List Archive: Cisco: NSP

OSPF's handling of secondary IP addresses on subinterfaces

Index | Next | Previous | View Threaded

fasterfourier at gmail

Apr 6, 2012, 7:29 PM

Post #1 of 7 (971 views)
Permalink

OSPF's handling of secondary IP addresses on subinterfaces

I have a FastEthernet interface with multiple 802.1Q based
subinterfaces. OSPF properly advertises routes for the subnets
directly defined on these subinterfaces. However if I add secondary IP
addresses to these subinterfaces, the subnets defined using the
secondary command show up in the local routing table as directly
connected, but OSPF doesn't advertise them to the rest of the area.
The OSPF process includes network statements that encompass both
primary and secondary IP addresses and they're all in the same area.

Are there any limitations I should know about with OSPF on
subinterfaces with secondary IP addresses?

IOS 12.4(25) on a 3745.

Thanks in advance.
_______________________________________________
cisco-nsp mailing list cisco-nsp [at] puck
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

wargo1 at gmail

Apr 7, 2012, 1:11 PM

Post #2 of 7 (929 views)
Permalink

Re: OSPF's handling of secondary IP addresses on subinterfaces [In reply to]

Hey Robert--

Yep, been there and had the same problem with the secondary IP address
on an interface (physical or subinterface). When I discovered this, I
researched it and found that OSPF on IOS does not put the subnet for the
subinterface in the database.

To get around that, you could probably a static route with next hop of
Null0 and tag then make a route map to advertise it.

cjw

Date: Fri, 6 Apr 2012 22:29:06 -0400
> From: Robert Johnson <fasterfourier [at] gmail>
> To: cisco-nsp [at] puck
> Subject: [c-nsp] OSPF's handling of secondary IP addresses on
> subinterfaces
> Message-ID:
> <CAOq=Mm=s8CQ1E+=HxNKqpdqohOP4w0fUdUTBiTG9eg7eWmr0LA [at] mail
> >
> Content-Type: text/plain; charset=ISO-8859-1
>
> I have a FastEthernet interface with multiple 802.1Q based
> subinterfaces. OSPF properly advertises routes for the subnets
> directly defined on these subinterfaces. However if I add secondary IP
> addresses to these subinterfaces, the subnets defined using the
> secondary command show up in the local routing table as directly
> connected, but OSPF doesn't advertise them to the rest of the area.
> The OSPF process includes network statements that encompass both
> primary and secondary IP addresses and they're all in the same area.
>
> Are there any limitations I should know about with OSPF on
> subinterfaces with secondary IP addresses?
>
> IOS 12.4(25) on a 3745.
>
> Thanks in advance.
>
>
>
_______________________________________________
cisco-nsp mailing list cisco-nsp [at] puck
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

gert at greenie

Apr 7, 2012, 1:50 PM

Post #3 of 7 (933 views)
Permalink

Re: OSPF's handling of secondary IP addresses on subinterfaces [In reply to]

Hi,

On Sat, Apr 07, 2012 at 03:11:10PM -0500, Christopher J. Wargaski wrote:
> To get around that, you could probably a static route with next hop of
> Null0 and tag then make a route map to advertise it.

This will only work if the static route could win against a connected
route - which it will never do (unless more specific, and in that case,
you have just null-routed your network, instead of connected it).

We just do "redistribute connected subnets" if we want connected
subnets in OSPF. Can be qualified by a route-map if needed.

(The original poster could just go away from using IPv4... IPv6 doesn't
have this funny idea about "primary" and "secondary" networks, and
neither does OSPFv3)

gert
--
USENET is *not* the non-clickable part of WWW!
//www.muc.de/~gert/
Gert Doering - Munich, Germany gert [at] greenie
fax: +49-89-35655025 gert [at] net

linkconnect at googlemail

Apr 7, 2012, 2:30 PM

Post #4 of 7 (923 views)
Permalink

Re: OSPF's handling of secondary IP addresses on subinterfaces [In reply to]

Sent from my iPad

On 7 Apr 2012, at 03:29, Robert Johnson <fasterfourier [at] gmail> wrote:

> I have a FastEthernet interface with multiple 802.1Q based
> subinterfaces. OSPF properly advertises routes for the subnets
> directly defined on these subinterfaces. However if I add secondary IP
> addresses to these subinterfaces, the subnets defined using the
> secondary command show up in the local routing table as directly
> connected, but OSPF doesn't advertise them to the rest of the area.
> The OSPF process includes network statements that encompass both
> primary and secondary IP addresses and they're all in the same area.
>
> Are there any limitations I should know about with OSPF on
> subinterfaces with secondary IP addresses?
>
> IOS 12.4(25) on a 3745.
>
> Thanks in advance.
> _______________________________________________
> cisco-nsp mailing list cisco-nsp [at] puck
>

I've seen this happen before. The same box announced the secondary subnet on one subinterface but did not on a another. We do have redist connected subnets enabled.

Regards

Wayne
_______________________________________________
cisco-nsp mailing list cisco-nsp [at] puck
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

fasterfourier at gmail

Apr 7, 2012, 9:10 PM

Post #5 of 7 (922 views)
Permalink

Re: OSPF's handling of secondary IP addresses on subinterfaces [In reply to]

Yes, I'm actually seeing one subinterface with a secondary address
have both the primary and secondary address subnets advertised into
OSPF and another have only the primary advertised. Very strange.
Suppose it's worth trying redistribute connected... thanks for the
tip.

Is this expected behavior or a bug?

On Sat, Apr 7, 2012 at 5:30 PM, Wayne Lee <linkconnect [at] googlemail> wrote:
>
>
> Sent from my iPad
>
> On 7 Apr 2012, at 03:29, Robert Johnson <fasterfourier [at] gmail> wrote:
>
>> I have a FastEthernet interface with multiple 802.1Q based
>> subinterfaces. OSPF properly advertises routes for the subnets
>> directly defined on these subinterfaces. However if I add secondary IP
>> addresses to these subinterfaces, the subnets defined using the
>> secondary command show up in the local routing table as directly
>> connected, but OSPF doesn't advertise them to the rest of the area.
>> The OSPF process includes network statements that encompass both
>> primary and secondary IP addresses and they're all in the same area.
>>
>> Are there any limitations I should know about with OSPF on
>> subinterfaces with secondary IP addresses?
>>
>> IOS 12.4(25) on a 3745.
>>
>> Thanks in advance.
>> _______________________________________________
>> cisco-nsp mailing list �cisco-nsp [at] puck
>>
>
> I've seen this happen before. The same box announced the secondary subnet on one subinterface but did not on a another. We do have redist connected subnets enabled.
>
> Regards
>
> Wayne
> _______________________________________________
> cisco-nsp mailing list �cisco-nsp [at] puck
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/

_______________________________________________
cisco-nsp mailing list cisco-nsp [at] puck
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

fasterfourier at gmail

Apr 9, 2012, 12:07 PM

Post #6 of 7 (922 views)
Permalink

Re: OSPF's handling of secondary IP addresses on subinterfaces [In reply to]

Okay, I believe I've found my problem.

My primary address statement had a typo that put it outside of the
OSPF network definitions. Apparently when the primary address on an
interface isn't picked up by a particular OSPF process, none of the
associated secondary addresses will be either.

Will test to confirm tonight.

On Sun, Apr 8, 2012 at 12:10 AM, Robert Johnson <fasterfourier [at] gmail> wrote:
> Yes, I'm actually seeing one subinterface with a secondary address
> have both the primary and secondary address subnets advertised into
> OSPF and another have only the primary advertised. Very strange.
> Suppose it's worth trying redistribute connected... thanks for the
> tip.
>
> Is this expected behavior or a bug?
>
>
> On Sat, Apr 7, 2012 at 5:30 PM, Wayne Lee <linkconnect [at] googlemail> wrote:
>>
>>
>> Sent from my iPad
>>
>> On 7 Apr 2012, at 03:29, Robert Johnson <fasterfourier [at] gmail> wrote:
>>
>>> I have a FastEthernet interface with multiple 802.1Q based
>>> subinterfaces. OSPF properly advertises routes for the subnets
>>> directly defined on these subinterfaces. However if I add secondary IP
>>> addresses to these subinterfaces, the subnets defined using the
>>> secondary command show up in the local routing table as directly
>>> connected, but OSPF doesn't advertise them to the rest of the area.
>>> The OSPF process includes network statements that encompass both
>>> primary and secondary IP addresses and they're all in the same area.
>>>
>>> Are there any limitations I should know about with OSPF on
>>> subinterfaces with secondary IP addresses?
>>>
>>> IOS 12.4(25) on a 3745.
>>>
>>> Thanks in advance.
>>> _______________________________________________
>>> cisco-nsp mailing list �cisco-nsp [at] puck
>>>
>>
>> I've seen this happen before. The same box announced the secondary subnet on one subinterface but did not on a another. We do have redist connected subnets enabled.
>>
>> Regards
>>
>> Wayne
>> _______________________________________________
>> cisco-nsp mailing list �cisco-nsp [at] puck
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/

_______________________________________________
cisco-nsp mailing list cisco-nsp [at] puck
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

gert at greenie

Apr 9, 2012, 12:17 PM

Post #7 of 7 (903 views)
Permalink

Re: OSPF's handling of secondary IP addresses on subinterfaces [In reply to]

Hi,

On Mon, Apr 09, 2012 at 03:07:12PM -0400, Robert Johnson wrote:
> Okay, I believe I've found my problem.
>
> My primary address statement had a typo that put it outside of the
> OSPF network definitions. Apparently when the primary address on an
> interface isn't picked up by a particular OSPF process, none of the
> associated secondary addresses will be either.

That's well-understood.

Area definitions in OSPF only look at the primary addresses (on Cisco IOS
devices, $J doesn't have the concept of a "primary" IPv4 address on
an interface at all).

gert
--
USENET is *not* the non-clickable part of WWW!
//www.muc.de/~gert/
Gert Doering - Munich, Germany gert [at] greenie
fax: +49-89-35655025 gert [at] net

Index | Next | Previous | View Threaded

Interested in having your list archived? Contact Gossamer Threads