I believe wireshark is trying to decode it as a different protocol which encapsulates IP in UDP. Try disabling it. On Oct 6, 2013 6:02 AM, "Chris Healy" cphealy@gmail.com wrote:

Alain,

I'm trying to perform a packet capture of the data transmitted by udp-sender using Wireshark and am running into the exact same problem encountered by Lance 8 years ago. (posting listed below.)

What information can I provide to help debug this issue?

Also, can you tell me if the multicast data sent includes a sequence number and does the receiver have the ability to filter out duplicate multicast frames?

Regards,

Chris

On Mon 8/15/2005 10:25 AM, Alain Knaff wrote:

...

*Try upgrading to a newer version of ethereal.*>**>*I just tried it here:

version 0.10.9 has the problem, but 0.10.12*>*works fine.*>**>*Or use tcpdump...* I have used the lattest version of udpcast (udpcast-20050226.tar.gz) as well as used ethereal version 0.10.12 and I still get the "Bogus IP header length (0, must be at least 20)" I will continue to try this on different Platforms with different network switches and see what I get.

-Lance _______________________________________________ Udpcast mailing list Udpcast@udpcast.linux.lu https://www.udpcast.linux.lu/cgi-bin/mailman/listinfo/udpcast

is it over Wi-Fi? There may be retransmissions, but they haven't caused me any issues. On Oct 6, 2013 10:06 AM, "Chris Healy" cphealy@gmail.com wrote:

Jayen,

Thank you. You were correct. I no longer have this problem. I also deduced that there is a sequence number in the multicast frame.

I'm digging into this utility as I currently have a scenario where I get a bunch of duplicate frames hitting the udp-receiver (network issue, not application issue) and the application does not gracefully handle it. I figured that the application did not have a sequence number. As it does, there must be some other problem. More to come.

Chris

On Sat, Oct 5, 2013 at 3:55 PM, Jayen jayenashar@gmail.com wrote:

...

I believe wireshark is trying to decode it as a different protocol which encapsulates IP in UDP. Try disabling it. On Oct 6, 2013 6:02 AM, "Chris Healy" cphealy@gmail.com wrote:

...

Alain,

I'm trying to perform a packet capture of the data transmitted by udp-sender using Wireshark and am running into the exact same problem encountered by Lance 8 years ago. (posting listed below.)

What information can I provide to help debug this issue?

Also, can you tell me if the multicast data sent includes a sequence number and does the receiver have the ability to filter out duplicate multicast frames?

Regards,

Chris

On Mon 8/15/2005 10:25 AM, Alain Knaff wrote:

...

*Try upgrading to a newer version of ethereal.*>**>*I just tried it

here: version 0.10.9 has the problem, but 0.10.12*>*works fine.*>**>*Or use tcpdump...*

I have used the lattest version of udpcast (udpcast-20050226.tar.gz) as well as used ethereal version 0.10.12 and I still get the "Bogus IP header length (0, must be at least 20)" I will continue to try this on different Platforms with different network switches and see what I get.

-Lance _______________________________________________ Udpcast mailing list Udpcast@udpcast.linux.lu https://www.udpcast.linux.lu/cgi-bin/mailman/listinfo/udpcast

Do you have a switching loop? (I guess if you found it, you'd have already addressed it.) On Oct 6, 2013 10:34 AM, "Chris Healy" cphealy@gmail.com wrote:

No, it's a wired network with a known flaw that leads to temporary packet storms of up to a second or so. (We are working on addressing this network flaw, but it seems to have exposed a vulnerability within this application.)

On Sat, Oct 5, 2013 at 4:27 PM, Jayen jayenashar@gmail.com wrote:

...

is it over Wi-Fi? There may be retransmissions, but they haven't caused me any issues. On Oct 6, 2013 10:06 AM, "Chris Healy" cphealy@gmail.com wrote:

...

Jayen,

Thank you. You were correct. I no longer have this problem. I also deduced that there is a sequence number in the multicast frame.

I'm digging into this utility as I currently have a scenario where I get a bunch of duplicate frames hitting the udp-receiver (network issue, not application issue) and the application does not gracefully handle it. I figured that the application did not have a sequence number. As it does, there must be some other problem. More to come.

Chris

On Sat, Oct 5, 2013 at 3:55 PM, Jayen jayenashar@gmail.com wrote:

...

I believe wireshark is trying to decode it as a different protocol which encapsulates IP in UDP. Try disabling it. On Oct 6, 2013 6:02 AM, "Chris Healy" cphealy@gmail.com wrote:

...

Alain,

I'm trying to perform a packet capture of the data transmitted by udp-sender using Wireshark and am running into the exact same problem encountered by Lance 8 years ago. (posting listed below.)

What information can I provide to help debug this issue?

Also, can you tell me if the multicast data sent includes a sequence number and does the receiver have the ability to filter out duplicate multicast frames?

Regards,

Chris

On Mon 8/15/2005 10:25 AM, Alain Knaff wrote:

...

*Try upgrading to a newer version of ethereal.*>**>*I just tried it

here: version 0.10.9 has the problem, but 0.10.12*>*works fine.*>**>*Or use tcpdump...*

I have used the lattest version of udpcast (udpcast-20050226.tar.gz) as well as used ethereal version 0.10.12 and I still get the "Bogus IP header length (0, must be at least 20)" I will continue to try this on different Platforms with different network switches and see what I get.

-Lance _______________________________________________ Udpcast mailing list Udpcast@udpcast.linux.lu https://www.udpcast.linux.lu/cgi-bin/mailman/listinfo/udpcast