Using IP Multicast Group range 239.0.0.(1-255)

[tmoffett]

Hi,

We've deployed IP Multicast in sparse mode on our network. We've defined a static RP for 239.0.0.x on every router. We are having problems with mcast in some sites that we are unable to figure out. Most of them involve 6500 switches in hybrid mode.

A TAC engineer told me that using the group range 239.0.0.x was probably causing my problems and didn't want to proceed with the troubleshooting until we changed the range to something that didn't overlap with 224.0.0.x at layer 2. Until mid-January we are unable to make those kinds of changes and I feel as if there are other issues outside of this.

Do you think this is really causing multicast to fail?

Any feedback would be greatly appreciated...

[subbarao.s]

Yes,

It sounds reasonable. Because 224.0.0.(1-255) is reserved group and all packets destined for this range are handled by switch CPU.

I hope you are aware of how Multicast address to MAC adress mapping is done. The lower 23 bits of IP address are mapped to sorresponding MAC address. So switch cant differentiate between reserved group (224.0.0.X) traffic and the group you are using (239.0.0.X) both mapes to identical MAC Address from 0x0100.5e00.0001 to 0x00100.5e00.00ff

If I am in your place, I certainly change the group adressing as a first step towards troubleshooting in this case.

regards,

Subba

[tmoffett]

Thank you for the reply! Yes, I am aware of the overlap, however I did overlook it at first...

We changed the group range to 239.1.0.x and things have improved somewhat.

I was reading a RFC1112 related document that lists all mcast addresses and their usage.

Am I okay with using 239.1.0.x? It appears to overlap with the "ST Multicast group" range, but I am unaware of why it would be in use in the Xerox network...

Is there any benefit in using 239.192.0.x instead?

Thanks,

Tim