We are running multiple Windows computers and they seem to have periodic network access problems.
We're running a Catalyst 3560. IOS is older version 12.2.25(SEA). We have multiple VLAN's setup on the switch with SVI's for each with an IP on different subnets assigned for each. We have a DHCP pool setup for VLAN 92 only.
Multiple client computers are setup for DHCP on VLAN 91 and they should be pulling an IP from a Windows DHCP server on that same VLAN. The Cisco switch is seeing the request and saying there's no address pool, which is correct since we are not running Cisco DHCP on that VLAN. I believe that the client machine then gets an IP from the Windows DHCP server, so the Cisco see's the client computers subnet change. However, it then sends a NAK to the client, which is logged in the Windows Event Log.
Shouldn't the switch ignore the request entirely and not participate in DHCP on this VLAN?
Oct 23 08:26:49: DHCPD: DHCPDISCOVER received from client 0100.234d.1a60.43 on interface Vlan91.
Oct 23 08:26:49: DHCPD: there is no address pool for 172.30.91.252.
Oct 23 08:26:53: DHCPD: DHCPDISCOVER received from client 0100.234d.1a60.43 on interface Vlan91.
Oct 23 08:26:53: DHCPD: there is no address pool for 172.30.91.252.
Oct 23 08:27:02: DHCPD: DHCPDISCOVER received from client 0100.234d.1a60.43 on interface Vlan91.
Oct 23 08:27:02: DHCPD: there is no address pool for 172.30.91.252.
Oct 23 08:27:02: DHCPD: DHCPREQUEST received from client 0100.234d.1a60.43.
Oct 23 08:28:30: DHCPD: DHCPREQUEST received from client 0100.234d.1a60.43.
Oct 23 08:28:30: DHCPD: client has moved to a new subnet.
Oct 23 08:28:30: DHCPD: Sending DHCPNAK to client 0100.234d.1a60.43.
