On 5/20/2010 1:12 PM, Sake Blok wrote:
On 20 mei 2010, at 18:45, Kevin Cullimore wrote:
On 5/20/2010 10:08 AM, Ronald Nutter wrote:
I am getting ready to make a change to the network at my company and need to verify what traffic is going over multicast.
I have been able to see existing multicast traffic by using a capture filter net 224.0.0.0/4. How can I see any multicast traffic that
Is on the network and isn�t using the normal multicast range ? Havent worked with multicast much, so just trying to be on the careful side.
In order to transmit& receive multicast traffic using non-standard destination addresses, you'd probably need to modify the protocol stacks of
-The participating hosts' operating system
-The operating systems of all IP-aware intermediate system the traffic would traverse
Sounds expensive, but some corporate interests go to great lengths in order to earn the "non-RFC-compliant" designation.
When faced with a task requiring a comprehensive collection of multicast packets seen on/through a given network/internetwork, I therefore worry far less about out-of-spec traffic such as you describe.
I totally agree, focus on the legitimate multicast traffic, but... if you're really concerned about non-compliant multicast traffic, you could search for it. Let's assume that there are no customized NICs in your company, then for the NIC to process a packet, it needs to be sent to:
1) the mac address of the NIC for unicast traffic... we're not interested in that traffic in your case
2) the broadcast address (ff:ff:ff:ff:ff:ff) for broadcast traffic... which we are also not interested in your case
3) a destination mac-address with the least significant bit set in the first octet for multicast traffic... that's what we can filter on :-)
So, to catch all non-RFC-compliant multicast traffic we can use the filter:
"ether multicast and not ether broadcast and not ip multicast"
Cheers,
Sake
Now, if you're concerned about L2 multicast, a likely source would be
Microsoft NLB hosts, and, IIRC, the above filter should do just fine
exposing that particular brand of miscreant. In general, eligible IP
multicast addresses should be a proper subset of ethernet multicast
addresses, although I'd be rather interested to hear if there are
exceptions.
___________________________________________________________________________
Sent via: Wireshark-users mailing list<wireshark-users@xxxxxxxxxxxxx>
Archives: http://www.wireshark.org/lists/wireshark-users
Unsubscribe: https://wireshark.org/mailman/options/wireshark-users
mailto:wireshark-users-request@xxxxxxxxxxxxx?subject=unsubscribe
