Hello
I've configure DHCP on the 3560 switches and noticed that it is causing a lot of havoc on my NAC.
When I display the binding table, I noticed a leading 01 attached to the MAC address and the same MAC has two entries.
See the following example:
SW004-01#sh ip dhcp binding | include fb
10.120.4.80 001c.259b.fb10 Jan 04 2012 01:58 PM Automatic
10.120.4.89 0100.1c25.9bfb.10 Jan 04 2012 01:59 PM Automatic
10.120.4.97 b8ac.6f80.e960 Dec 27 2011 02:13 AM Automatic
10.120.4.98 01b8.ac6f.80e9.60 Dec 30 2011 10:11 AM Automatic
10.120.4.137 0024.e843.f644 Jan 04 2012 09:03 AM Automatic
10.120.4.129 0100.24e8.43f6.44 Jan 04 2012 09:03 AM Automatic
SW004-01#
I can not PING IPs mapped with the 12 digts MACs, but can PING IPs mapped to 14digts MACs (ARP still shows entries to 12 digts MACs).
What is happenting, the NAC receives a packet, 10.120.4.80, tries to connect to that IP to validate the host, but fails to connect as the host now has a new IP address for 10.120.4.89. It is causing massive false-positives on the NAC side.
Why is this happeing on the switch?
Two IP addresses are being leased to same host, and this will quickly exhaust my pool.
Is there a fix?
Thanks