Can you share the link, for future reference?

On Wed, Jan 15, 2020 at 2:15 PM Juanjo Martin Carrascosa <jua...@rti.com>
wrote:

> Found the presentation, this is fantastic.
>
> Thanks!
>
> On Wed, Jan 15, 2020 at 12:58 PM Anders Broman via Wireshark-dev <
> wireshark-dev@wireshark.org> wrote:
>
>> Hi,
>>
>> In the frame layer there is the “Encapsulation type” the way the pcap
>> format works this indicates how the data following should be interpreted.
>>
>> Linktype/encapsulation type is defined at
>> https://www.tcpdump.org/linktypes.html so one thing you could do is to
>> design your own DLT
>>
>> Add whatever meta data you want and then add the RTPS data at some
>> position in that structure. You can try this out by using the USER DLTs in
>> Wireshark
>>
>> Or if you do not want to request a DLT from tcpdump ( DO NOT USE ONE
>> WITHOUT ALLOCATING IT) you could use the Exported PDU DLT defined by
>> Wireshark
>>
>> Epan/exported_pdu.h contains some information. If you would require new
>> elements for meta information that could be discussed. Adding RTSP data in
>>
>> An exported_pdu DLT frame should be simple as an initial test. I think
>> there is a sharkfest presentation on how to add any protocol data in an
>> exported pdu frame with text2pcap.
>>
>> Regards
>>
>> Anders
>>
>>
>>
>> *From:* Wireshark-dev <wireshark-dev-boun...@wireshark.org> *On Behalf
>> Of *Juanjo Martin Carrascosa
>> *Sent:* den 15 januari 2020 12:24
>> *To:* Developer support list for Wireshark <wireshark-dev@wireshark.org>
>> *Subject:* Re: [Wireshark-dev] Bypassing the first layer
>>
>>
>>
>> Hi Dario,
>>
>>
>>
>> Could you elaborate on this? I get the idea but my knowledge about the
>> how is very basic.
>>
>>
>>
>> How can this help me achieve what I need?
>>
>>
>>
>> Thanks for the help.
>>
>> Juanjo Martin
>>
>>
>>
>> On Tue, Jan 14, 2020 at 4:48 PM Dario Lombardo <lom...@gmail.com> wrote:
>>
>> You can use export_pdu. This is a layer with multiple attributes (see
>> epan/exported_pdu.h), but the only really needed is the proto_name, that
>> allows you to directly call a dissector by its name.
>>
>> You will have frames with this stack: FRAME/EXPORT PDU/RTPS, that is
>> properly dissected by wireshark.
>>
>>
>>
>> On Tue, Jan 14, 2020 at 4:33 PM Juanjo Martin Carrascosa <jua...@rti.com>
>> wrote:
>>
>> Hi everyone,
>>
>>
>>
>> RTPS is a protocol already supported by Wireshark. I have been helping
>> maintain that protocol these past years. It sits on top of TCP and UDP, as
>> well as some other transports that are not network ones (Shared Memory
>> typically with a proprietary implementation).
>>
>>
>>
>> We are currently working on implementing a new logging mechanism for our
>> product, mainly to address the Shared Memory scenario but it can also be
>> really useful when RTPS is used on top of network protocols.
>>
>>
>>
>> Problem: We can log the RTPS layer but we don't have some information in
>> our middleware like the Frame or Ethernet information, so we need to make
>> it up. This is ugly, so I am trying to avoid that:
>>
>>
>>
>> How can I register a protocol so it is picked up instead of the Frame
>> layer? This is, I want to create a new protocol that detects that the
>> information we generate is for that protocol and the Frame protocol is not
>> called, but the new protocol I am creating. This new protocol will then
>> call the RTPS protocol to dissect the payload I want to display. We are
>> planning to also add some information to this custom protocol, that's why I
>> want it to be called first.
>>
>>
>>
>> Note: I just came up with this solution, but if you have a different
>> solution for this, please let me know.
>>
>>
>>
>> Thanks,
>>
>> Juanjo Martin
>>
>>
>> ___________________________________________________________________________
>> Sent via:    Wireshark-dev mailing list <wireshark-dev@wireshark.org>
>> Archives:    https://www.wireshark.org/lists/wireshark-dev
>> Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
>>              mailto:wireshark-dev-requ...@wireshark.org
>> ?subject=unsubscribe
>>
>>
>>
>>
>> --
>>
>> Naima is online.
>>
>>
>> ___________________________________________________________________________
>> Sent via:    Wireshark-dev mailing list <wireshark-dev@wireshark.org>
>> Archives:    https://www.wireshark.org/lists/wireshark-dev
>> Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
>>              mailto:wireshark-dev-requ...@wireshark.org
>> ?subject=unsubscribe
>>
>>
>>
>>
>> --
>>
>>
>>
>> Juanjo Martin
>>
>> Principal Application Engineer
>>
>> EMEA Services Lead @ Professional Services Group
>>
>> Office: +34 958 27 88 62
>>
>> jua...@rti.com
>>
>> www.rti.com
>> <https://protect2.fireeye.com/v1/url?k=ef5c0e3a-b3d6db2c-ef5c4ea1-862f14a9365e-a65581b6c9ebf84e&q=1&e=75a63928-4ca4-405f-a42a-975bcffca831&u=http%3A%2F%2Fwww.rti.com%2F>
>>
>>
>>
>> ___________________________________________________________________________
>> Sent via:    Wireshark-dev mailing list <wireshark-dev@wireshark.org>
>> Archives:    https://www.wireshark.org/lists/wireshark-dev
>> Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
>>              mailto:wireshark-dev-requ...@wireshark.org
>> ?subject=unsubscribe
>
>
>
> --
>
> Juanjo Martin
> Principal Application Engineer
> EMEA Services Lead @ Professional Services Group
> Office: +34 958 27 88 62
> jua...@rti.com
> www.rti.com
>
>
> ___________________________________________________________________________
> Sent via:    Wireshark-dev mailing list <wireshark-dev@wireshark.org>
> Archives:    https://www.wireshark.org/lists/wireshark-dev
> Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
>              mailto:wireshark-dev-requ...@wireshark.org
> ?subject=unsubscribe



-- 

Naima is online.
___________________________________________________________________________
Sent via:    Wireshark-dev mailing list <wireshark-dev@wireshark.org>
Archives:    https://www.wireshark.org/lists/wireshark-dev
Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
             mailto:wireshark-dev-requ...@wireshark.org?subject=unsubscribe

Reply via email to