Heh, I just got the following note from Bill Fenner, re: the idmr
minutes from Oslo. One of the notes is:
IGMP
v2 spec - RFC 2236
Status: time to progress it to Draft Standard. Needs an
update to incorporate Erik Nordmark's comments on
the IPv6 MLD spec (most of which apply to the IGMPv2
spec as well). Also should mention that the common
practice of not reporting any group in 224.0.0/24
is not valid (only 224.0.0.1 is excluded).
At 12:43 PM 7/20/99 -0400, Eric S. Johnson wrote:
>
>I have some ethernet switches which do igmp snooping to determine
>what ports to deliver multicast traffic. This works pretty well with
>most systems, but with the linux boxs (2.2.X) there are problems.
>
>The linux boxes do not send IGMP packets when the host joins
>any "local" multicast group. Local being defined as 224.0.0.X.
>
>So for example, it is difficult to get OSPF on linux to work when
>the linux box is attached to these IGMP snooping switches. The linux
>boxes don't report membership in 224.0.0.[45]. The switches then don't
>forward packets addressed to those addresses to the linux box.
>
>Sooo...
>
>Is 224.0.0.X a "LOCAL net only" set of addresses? Is this defined
>somewhere? Is there a RFC (1700?) I can wave in the switch vendors
>face?
>
>Conversly, if I i just change the definition of LOCAL_MCAST in <linux/in.h>
>to
>
>#define LOCAL_MCAST(x) (((x) & htonl(0xFFFFFFFE)) == htonl(0xE0000000))
>
>IE make 224.0.0.1 the only "local" group, will that break anything else?
>
>E
>
>
>
>
>-
>To unsubscribe from this list: send the line "unsubscribe linux-net" in
>the body of a message to [EMAIL PROTECTED]
>
>
-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to [EMAIL PROTECTED]
