Also, please see
http://www.cisco.com/en/US/tech/tk648/tk361/technologies_tech_note09186a0080133ddd.shtml

This clearly outlines that when going from inside --> outside routing
should be performed first. Seems that is true, except in the specific case
I posted about (outside NAT)


On Mon, Apr 22, 2013 at 4:44 PM, Joe Astorino <[email protected]>wrote:

> Hi Kevin,
>
> Thanks for the reply, but your example is talking about what would happen
> on an outside --> inside flow.  In that case what you are saying makes
> sense.  However, when going from inside to outside I believe the rules
> generally are different.
>
> I think that for basic static NAT from inside --> outside indeed routing
> is performed first.  You can see this in the output of a packet-trace and
> most documentation I've seen shows route lookup happening first when going
> inside --> outside.
>
> It is only in this specific case (inside to outside flow with outside
> static NAT) that I am puzzled.
>
>
> On Mon, Apr 22, 2013 at 4:38 PM, Kevin Sheahan <[email protected]>wrote:
>
>> Hi Joe,
>>
>> For ASA code in both pre 8.3 and 8.3+, routing will be post-NAT. I don't
>> know the exact reasoning for this but I can speculate that it is to both
>> route based on 'real ip' (simplifies routing table) as well as to allow for
>> simpler implementation of 'route-lookup' NAT function. The order of
>> operations changes between pre 8.3 and 8.3+ between the "Access-Control"
>> and "NAT" steps (ACL done before nat in pre-8.3, NAT done before ACL in
>> 8.3+).
>>
>> If routing were to happen pre-NAT, than consider the following example:
>>
>>
>>    - You have 12.232.232.0/24 address space.
>>    - Your outside interface is assigned 12.232.232.80.
>>    - The whole 12.232.232.0/24 network is then directly connected to the
>>    outside interface. So any incoming packet being routed first would want to
>>    hairpin and not reach its correct destination. Additionally, RPF check
>>    would fail for NAT (unless NAT specifies outside,outside) so the packet
>>    would be dropped.
>>    - Because, in reality, routing happens after NAT – An incoming packet
>>    can be un-nat'ed to, for example, DMZ address 192.168.232.x and routed
>>    accordingly.
>>
>> Hope I was helpful.
>>
>> Good studies,
>>
>> Kevin Sheahan
>>
>> From: Joe Astorino <[email protected]>
>> Date: Monday, April 22, 2013 3:50 PM
>> To: OSL Security <[email protected]>
>> Subject: [OSL | CCIE_Security] 8.2 static outside NAT
>>
>> I could really use some clarification here. Here is my setup
>>
>> ASA running 8.2 code.  nat-control is not enforced.  Requirement is that
>> traffic destined to 192.168.10.241 on the inside will have the destination
>> translated to 10.12.20.56 on the outside.  Conversely, traffic sourced from
>> 10.12.20.56 on the outside will have it's source translated to
>> 192.168.10.241 on the inside.
>>
>> My solution
>>
>> static (outside,inside) 192.168.10.241 10.12.20.56 netmask 255.255.255.255
>>
>>
>> Now, I assumed going from inside --> outside routing happens first.  So,
>> I added a route like so
>> route (outside) 192.168.10.241 255.255.255.255 outside_next_hop
>>
>> This failed to work.  Only when I add a static route pointing outside for
>> the REAL address does this work.  This is baffling me.
>>
>> Also, when running packet-tracer the first step is UN-NAT which I've
>> never heard of before and can't find much information on.  Can anybody
>> explain why routing is happening POST nat here???
>> --
>> Regards,
>>
>> Joe Astorino
>> CCIE #24347
>> http://astorinonetworks.com
>>
>> "He not busy being born is busy dying" - Dylan
>> _______________________________________________ For more information
>> regarding industry leading CCIE Lab training, please visit
>> www.ipexpert.com Are you a CCNP or CCIE and looking for a job? Check out
>> www.PlatinumPlacement.com
>>
>
>
>
> --
> Regards,
>
> Joe Astorino
> CCIE #24347
> http://astorinonetworks.com
>
> "He not busy being born is busy dying" - Dylan
>



-- 
Regards,

Joe Astorino
CCIE #24347
http://astorinonetworks.com

"He not busy being born is busy dying" - Dylan
_______________________________________________
For more information regarding industry leading CCIE Lab training, please visit 
www.ipexpert.com

Are you a CCNP or CCIE and looking for a job? Check out 
www.PlatinumPlacement.com

Reply via email to