ANNOUNCEMENT: Live Wireshark University & Allegro Packets online APAC Wireshark Training Session
July 17th, 2024 | 10:00am-11:55am SGT (UTC+8) | Online

Wireshark-dev: Re: [Wireshark-dev] On building better statistics

From: João Valverde <j@xxxxxx>
Date: Tue, 15 Feb 2022 20:28:01 +0000


On 15/02/22 20:01, Jaap Keuter wrote:


On 15 Feb 2022, at 13:20, João Valverde <j@xxxxxx> wrote:

And also, it's not really correct to include IP inside ICMP as IP bytes, but that's another issue entirely.

Looking at the Protocol Hierarchy statistics, only the ‘top level’ protocols are counted. So, an IP header in a ICMP packet don’t get added.

I got that from the WSUG: https://www.wireshark.org/docs/wsug_html/#ChStatHierarchy.

"A single packet can contain the same protocol more than once. In this case, the protocol is counted more than once. For example ICMP replies and many tunneling protocols will carry more than one IP header."

Tunneling such as IP over IP I don't really see a use case where counting more than once would be useful, but it's not technically incorrect (unlike ICMP).

You can see that with IPv6, there you can choose to have the IPv6 extension headers under the root tree rather than under the IPv6 packet. If you put them under the root, they’ll show up in the Protocol Hierarchy statistics, otherwise not.

Regards,
Jaap


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