Why do you want to change the interface naming?
However we cannot change the "zc:” prefix, it identifies the module.

Alfredo

> On 29 Jan 2016, at 05:06, Madhu R <[email protected]> wrote:
> 
> HI Alfredo,
> Thanks for your help ,I had implemented the customized distribution function 
> ,is it possible to change zc:99@0 to zc99@0 when we are running the instance 
> as follows
> ./pfcount -i zc:99@0    to     ./pfcount -i zc99@0
> 
> Regards
> Madhu
> 
> 
> From: [email protected] 
> <[email protected]> on behalf of Alfredo Cardigliano 
> <[email protected]>
> Sent: Thursday, January 28, 2016 2:14 PM
> To: [email protected]
> Subject: Re: [Ntop-misc] PF_RING Load balancer
> 
> The distribution function should return the queue id, if you have 2 pfcount 
> (i.e. 2 queues) it should return 0 or 1.
> 
> Alfredo
> 
>> On 28 Jan 2016, at 01:48, Madhu R <[email protected] 
>> <mailto:[email protected]>> wrote:
>> 
>> How to send packet to a particular instance from the custom load 
>> distribution function (for example when we are running two pfcount 
>> instances) what will be the return value in the distribution function.
>> Regards
>> Madhu
>> 
>> 
>> 
>> On Wed, Jan 27, 2016 at 7:57 AM -0800, "Alfredo Cardigliano" 
>> <[email protected] <mailto:[email protected]>> wrote:
>> 
>> 
>>> On 27 Jan 2016, at 15:07, Madhu R <[email protected] 
>>> <mailto:[email protected]>> wrote:
>>> 
>>> Hi Alfredo,
>>> 
>>> As per your suggestion am trying to implement load balancing  custom 
>>> distribution function based on specific protocol ID in zbalance_ipc.c
>>> 
>>> 1)how to get the protocol ID from the buffer am not able to find out the 
>>> structure which is having the protocol ID
>> 
>> ZC does not parse the packet by default, you can use the pfring_parse_pkt() 
>> API, example:
>> 
>> struct pfring_pkthdr h;
>> memset(&h, 0, sizeof(h));
>> h.len = h.caplen = pkt_handle->len;
>> pfring_parse_pkt(pfring_zc_pkt_buff_data(pkt_handle, in_queue), &h, 4, 0, 0);
>> protocol = h.extended_hdr.parsed_pkt.l3_proto;
>> 
>>> 2)what is the size of the buffer used for capturing and storing the packets
>> 
>> Is the second parameter of pfring_zc_create_cluster (e.g. in zbalance_ipc.c 
>> is max_packet_len(devices[0]))
>> 
>>> 3)in which byte space in buffer i can find the protocol ID of the packets
>> 
>> As 1 (you need to parse the packet)
>> 
>>> 4)what are the elements in the structure  pfring_zc_queue
>> 
>> struct pfring_zc_queue is an internal structure, you do not need to access 
>> it.
>> 
>>> int32_t ip_distribution_func(pfring_zc_pkt_buff *pkt_handle, 
>>> pfring_zc_queue *in_queue, void *user)
>>>  5)what is void *user in the above function
>> 
>> The “user_data" parameter in pfring_zc_run_balancer() is passed to 
>> ip_distribution_func() as “user”, please refer to the doxygen documentation.
>> 
>> Regards
>> Alfredo
>> 
>>> Regards
>>> Madhu
>>> From: [email protected] 
>>> <mailto:[email protected]> 
>>> <[email protected] 
>>> <mailto:[email protected]>> on behalf of Alfredo 
>>> Cardigliano <[email protected] <mailto:[email protected]>>
>>> Sent: Friday, January 22, 2016 4:05 PM
>>> To: [email protected] <mailto:[email protected]>
>>> Subject: Re: [Ntop-misc] PF_RING Load balancer
>>> 
>>> 
>>>> On 21 Jan 2016, at 04:54, Madhu R <[email protected] 
>>>> <mailto:[email protected]>> wrote:
>>>> 
>>>> 
>>>> Hi ,
>>>> I have a doubt regarding working of PF_RING as a load balancer with snort  
>>>> as IDS/IPS ,
>>>> 
>>>> 1) Is PF_RING is able to do load balancing depending on the type of 
>>>> packets (ex TCP, UDP, ICMP etc) to snort?
>>> 
>>> If you are using vanilla drivers you have a few options in the daq module 
>>> (--daq-var clustermode=<mode>), setting mode to:
>>>    - 2 for 2-tuple flow
>>>    - 4 for 4-tuple flow
>>>    - 5 for 5-tuple flow
>>>    - 6 for 6-tuple flow
>>> 
>>> If you are using PF_RING ZC, you can use hw RSS (4-tuple) or zbalance_ipc 
>>> -m 1 which is an IP-based hash.
>>> 
>>> If you need something more specific (e.g. protocol only), it is possible to 
>>> define a custom distribution function in zbalance_ipc.c
>>> 
>>>> 2) In case if its not possible is there a way to do load balancing 
>>>> depending on type of packet ?
>>>> 3) Which example i need to consider for doing it or do i need to do 
>>>> modification in the source code for the packet classification ?
>>> 
>>> As above.
>>> 
>>> Regards
>>> Alfredo
>>> 
>>>> 
>>>> Thanks and regards
>>>> Madhu
>>>> _______________________________________________
>>>> Ntop-misc mailing list
>>>> [email protected] <mailto:[email protected]>
>>>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc 
>>>> <http://listgateway.unipi.it/mailman/listinfo/ntop-misc>
>>> _______________________________________________
>>> Ntop-misc mailing list
>>> [email protected] <mailto:[email protected]>
>>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc 
>>> <http://listgateway.unipi.it/mailman/listinfo/ntop-misc>
>> _______________________________________________
>> Ntop-misc mailing list
>> [email protected] <mailto:[email protected]>
>> http://listgateway.unipi.it/mailman/listinfo/ntop-misc
> 
> _______________________________________________
> Ntop-misc mailing list
> [email protected] <mailto:[email protected]>
> http://listgateway.unipi.it/mailman/listinfo/ntop-misc 
> <http://listgateway.unipi.it/mailman/listinfo/ntop-misc>

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

_______________________________________________
Ntop-misc mailing list
[email protected]
http://listgateway.unipi.it/mailman/listinfo/ntop-misc

Reply via email to