[Click] Accessing the ingress interface of a packet

[Eddie Kohler]

As you work more with Click you'll get a better feeling for what it is 
and isn't best for.  The modular architecture is precisely designed for 
complex operations that are the composite result of several simple 
steps.  Your example should be implementable in packet oriented 
dataflow.  GetIPAddress will fetch the address being resolved in step 
(1) and store it in an annotation, RadixIPLookup will do a lookup & 
split the packets based on the lookup, and you will route packets to the 
correct ARPQuerier based on the output of RadixIPLookup.

Eddie


On 1/9/11 2:09 PM, Philip Prindeville wrote:
> That's what I thought.
>
> I was wondering about the scenario where I might want to do the following:
>
> (1) grovel inside an ARP-Reply packet for the address being resolved;
> (2) do an RadixIPLookup() on that address
> (3) locate the appropriate ARPQuerier instance based on that lookup;
> (4) forward that ARP-Reply to the appropriate ARPQuerier instance.
>
> Being purely packet-oriented is somewhat limiting, since more complex
> operations that modify the packet flow might need to be the composite
> result of several primitive steps, as above.
>
>
>
> On 1/9/11 10:55 AM, Eddie Kohler wrote:
>> Hi Philip,
>>
>> I read the rest of this thread but just to respond to the top level
>> question.
>>
>> Elements don't really "RETURN" anything. They are parts of a
>> packet-oriented dataflow.
>>
>> The ingress interface of a packet is accessible from within C++,
>> however. For the full interface to Packet:
>>
>> http://read.cs.ucla.edu/click/doxygen/classPacket.html
>>
>> You want Packet::device_anno().
>>
>> Eddie
>>
>>
>> On 1/1/11 12:26 PM, Philip Prindeville wrote:
>>> (Forgive the newbie questions... I'm having to learn Click in a hurry
>>> as I ramp up on a project.)
>>>
>>> Is there a built-in element that returns the ingress interface of a
>>> packet?
>>>
>>> I'm trying to write an elementclass in Click that modifies a packet
>>> and then sends it back out the same interface it came in on, but
>>> don't want to have one instance of it per-interface that it is used on.
>>>
>>> Thanks.
>>>
>>> -Philip
>>>
>>> _______________________________________________
>>> click mailing list
>>> click at amsterdam.lcs.mit.edu
>>> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
>