Re: Switches and Multicast doubt

 You're thinking L3 multicast though if you're only consideration is IGMP
snooping!

A frame destined for FFFF.FFFF.FFFF is a broadcast. (all 1's in binary).
Specifically THOSE will match the broadcast counter.

Any frame where the first byte (two hex characters) is xxxx xxx1 is a
multicast (hex ending in 1, 3, 5, 7, 9, B, D or F). From a counter
standpoint, it would be a matter of semantics whether the multicast count
included the broadcast or not. Personally, I wouldn't program it that
way, but that's just me. I've never really paid attention to it to
determine that! If you ever see your broadcast counter higher than your
multicast counter, you know they're separated!

All broadcasts are indeed multicasts (group messages). All multicasts
are NOT broadcasts.

IGMP snooping though is a specific subset used for forwarding decisions.
Much like a unicast MAC has a specific port assigned to it, based on
information seen in IGMP, multicast MAC entries will only have a few
ports associated with it other than the flooding which happens by
default.

Scott Morris, CCIEx4 (R&S/ISP-Dial/Security/Service Provider) #4713,

CCDE #2009::D, CCNP-Voice, JNCIE-SP #153, JNCIE-ER #102, CISSP, et al.

CCSI #21903, JNCI-SP, JNCI-ER

swm_at_emanon.com

Knowledge is power.

Power corrupts.

Study hard and be Eeeeviiiil......

On 9/29/11 3:05 AM, CCIE KID wrote:

  Still i find two different interface counters for broadcast and
  multicast. How does the switch internally differentiate the multicast
  and broadcast. If IGMP snooping is enabled , the switch the snoop the
  packets going from router to the host . And it updates in its MAC
  table about the multicast MAC address and the corresponding
  interface. How does the switch look into the frame and update in its
  MAC table. Can anyone explain me the procedure. I am confused?THank
  you
  On Wed, Sep 28, 2011 at 10:13 PM, Scott Morris <swm_at_emanon.com>
  wrote:

    At Layer2, the difference is subtle. At Layer3, it's different!
     As
    others pointed out, 01:00:5E will lead every IP multicast packet
    at
    Layer2 (it's all about the handoff to higher functions in the
    switch).

    Any other Layer2 "group" address which does not have specific
    code
    attached to it will be replicated to all devices in that
    broadcast domain.

    There are many group addresses though, that DO have other code to
    intercept... Spanning Tree is an L2 multicast address.... CDP
    is...
    Just to name a couple!

    It's all about awareness though. if the switch isn't aware, it
    will
    flood the frame. The PC then is presumed to be aware of whether
    it's
    part of that group or not.

    HTH,

    Scott

    *Scott Morris*, CCIE/x4/ (R&S/ISP-Dial/Security/Service Provider)
    #4713,

    CCDE #2009::D, CCNP-Voice, JNCIE-SP #153, JNCIE-ER #102, CISSP,
    et al.

    CCSI #21903, JNCI-SP, JNCI-ER

    swm_at_emanon.com

    Knowledge is power.

    Power corrupts.

    Study hard and be Eeeeviiiil......

    On 9/28/11 11:10 AM, CCIE KID wrote:
> Hi fellas,
>
> I have a million dollar doubt. Whether switches understand
    multicast or not.
> Its all about the so called I/G bit setting in the MAC frame.
    If the I bit
> is set to 1 then its a multicast address . But unfortunately
    the broadcast
> address also has the I bit set to 1 . So how does a switch
    differentiate
> multicast and broadcast. Is there any method used in cisco
    switches to
> differentiate. Can anyone help me out?

    Blogs and organic groups at http://www.ccie.net

    _______________________________________________________________________
    Subscription information may be found at:
    http://www.groupstudy.com/list/CCIELab.html

  --
  With Warmest Regards,

  CCIE KID
  IN PURSUIT OF CCIE

Blogs and organic groups at http://www.ccie.net
Received on Thu Sep 29 2011 - 07:48:15 ART