Building network automation solutions

9 module online course

Start now!

DHCP Client Address Change Detector

In a previous post I’ve described how useless DHCP logging is when you try to detect change in DHCP-assigned IP address. Fortunately the removal of the old IP address (triggered by the DHCPNAK server response) and configuration of the new IP address (sent in the DHCPACK response) triggers a change in the IP routing table that can be detected with the IP routing table event detector introduced in EEM 3.0 (available from Cisco IOS release 12.4(22)T).

The following text written by Ivan Pepelnjak in 2009 was originally published on CT3 wiki. That web site became unreachable in early 2019. We retrieved the original text from the Internet Archive, cleaned it up, updated it with recent information if necessary, and republished it on blog on December 5, 2020
Change in the client IP address assigned by a DHCP server to a Cisco router is hard to detect due to lack of consistent DHCP-related syslog message. Cisco IOS generates a DHCP-6-ADDRESS_ASSIGN syslog message when an interface configured with ip address dhcp acquires its initial address through DHCP, but not when the address is subsequently changes due to a DHCP NAK response from the DHCP server.
Change in an interface IP address always generates a change in the IP routing table: the connected route to the subnet of the old IP address is removed and new connected route is inserted in the IP routing table. The change in the IP routing table can be detected with the IP routing event detector introduced in EEM 3.0.

The following EEM applet requires the IP routing event detector, programming logic and regular expression support available in Embedded Event Manager 3.0 (first released with Cisco IOS release 12.4(22)T).

Applet source code
event manager applet DHCPAddressChange
 event routing network type add protocol connected ge 1
 action 100 cli command "show ip interface brief | include $_routing_lastinterface"
 action 110 regexp "YES\s+DHCP" "$_cli_result"
 action 200 if $_regexp_result eq 1
 action 220  set ipaddress "unknown"
 action 230  regexp " ([0-9.]+) " "$_cli_result" match ipaddress
 action 240  info type routername
 action 250  mail server "$_mail_smtp" to "$_mail_rcpt" from "[email protected]$_mail_domain" →
             subject "DHCP address on $_routing_lastinterface changed to $ipaddress" →
             body "\n$_cli_result"
 action 299 end 

The event routing network type add protocol connected event detector detects all additions of connected routes (the mask indicates we want to catch all changes regardless of the actual IP prefix).

Connected routes could change after interface flaps or manual configuration changes. The show ip interface brief command is thus used to inspect the IP address allocation method of the affected interface (the interface name is available in the $_routing_lastinterface EEM variable).

The presence of “YES\s+DHCP” pattern in the show ip interface brief command output indicates the interface is operational and got its IP address via DHCP. In this case, the actual IP address is extracted from the show ip interface brief command output and sent in an e-mail to the network operator.

Additional configuration

The EEM applet expects several EEM environmental variables defining your SMTP environment. A sample configuration is included in the following printout (you could also replace these variables with hardcoded values in the EEM applet).

event manager environment _mail_smtp
event manager environment _mail_domain
event manager environment _mail_rcpt [email protected]


  1. I have a DHCP server installed in a Windows 2003 server that contains several
    Address pool assigned to each VLAN. in my switch (Cisco 4506), I configured each SVI interface with an DHCP-relay to send DHCP request to my server.actually its all going well
    But my point if a user changes the VLAN it keeps
    the same address given in the VLAN, but not a new address that corresponds to the new VLAN.
    How to solve this problem????
  2. First, the client has to detect it's on another VLAN; its LAN interface must flap.

    Second, once it sends out its DHCP request asking the DHCP server to reconfirm its address, the switch has to be willing to propagate it.

    Last, the DHCP server must be "authoritative" (ISC's terminology, not mine), it must reply with a NAK, thus forcing the client to ask for another address.
Add comment