Valid mac address

Feb 08, 2006 10 Replies

Is 00 00 00 00 00 00 a valid mac address which could be subjected to learning and forwarding ?



Thanks in advance Srinivas



No.

But that doesn't mean that misconfigured or malfunctioning equipment will never generate it.

There is nothing syntactically *invalid* about that address, either as a SA or a DA. However, since the OUI 00-00-00 has not been assigned to any equipment manufacturer, no device should be using it. That said, it is possible that a device may emit frames to/from that address under fault conditions (e.g., the device is broken in some fashion), so a *receiver* (e.g., a bridge or end station) should be prepared to deal with frames sent to/from the address, and not produce undesirable results.

-- Rich Seifert Networks and Communications Consulting 21885 Bear Creek Way (408) 395-5700 Los Gatos, CA 95033 (408) 228-0803 FAX

Send replies to: usenet at richseifert dot com

Didn't some antique machines, I dimly remember Suns perhaps, use the all 0s MAC address as a broadcast address?

Bert

Didn't some antique machines, I dimly remember Suns perhaps, use the all 0s MAC address as a broadcast address?

Bert

Also, some higher layers assume that the all-zero address is a place holder for an unknown MAC address.

Hmmm, interesting,

formatting link
that 00-00-00 is allocated to Xerox, and
formatting link
appears to imply that only 01-80-C2-00-00-00 to 01-80-C2-00-00-0F are reserved as not-to-be-forwarded by MAC bridges.

Albert Manfredi wrote: (snip)

(snip)

SunOS by default uses the 0 broadcast IP address. That is, all 0 for the host part instead of the more usual all ones. Manually setting the broadcast address on ifconfig fixes it.

As I understand it, using this on a normal network can generate broadcast storms.

-- glen

I believe that that "assignment" is simply a placeholder; Xerox "assigned" it to itself (and never used it) because RFC1042 specifies the use of the 00-00-00 OUI for SNAP encapsulation of Ethernet Type fields. At the time of RFC1042 (the early 1980s) there was no better "authority" to assign it to.

That is correct; this address block includes the multicasts used for the Spanning Tree Protocol and PAUSE flow control.

-- Rich Seifert Networks and Communications Consulting 21885 Bear Creek Way (408) 395-5700 Los Gatos, CA 95033 (408) 228-0803 FAX

Send replies to: usenet at richseifert dot com

Minor nit - as of 802.1ad, some of these addresses are specified as non-reserved for provider bridges - they get forwarded just like regular multicast frames in the S-VLAN component of provider bridges.

Anoop

I once knew a net that had such a device. As there was only one, the decision was to just leave it alone. One possibility is an address stored in EEPROM that gets cleared and no new address set.

-- glen

Occasionally a vendor will forget to program the MAC address for a card (or run of cards) and then it'll sometimes come up as all zeros.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required