cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
3165
Views
0
Helpful
8
Replies

Interface reset on DHCP renewal?

dlandriscina
Level 1
Level 1

As anyone experienced this problem:

Im running POE for Cisco 504G phones with a PC attached to the LAN switch port on the phone.  DHCP is being served through a Cisco 1800 series router.

It seems that the interface will shutdown and come right back on at DHCP renewal obviously resetting the phone and causing the PC client to disconnect.  The only reason I noticed this is because it is right at the same time the lease is expiring.  Both the phone and the PC are on the same VLAN.

Thanks

Domenick

8 Replies 8

michael.leblanc
Level 4
Level 4

Domenick:

The following link makes reference to the issue (Expired DHCP lease bounces the interface):

http://blog.ioshints.info/2009/09/expired-dhcp-lease-bounces-interface.html

... which confirms that you are not the only one to witness this behaviour.

Additional information (that may be of interest to you) on the behaviour of the Cisco IOS DHCP client, and server, may be found at:

http://wiki.nil.com/DHCP_client_address_change

Best Regards,

Mike

Michael, thanks for responding to the thread!  Is there a fix for this?

Domenick

Domenick:

Don't know.

But the author of the article seems open to receiving communications. You might contact him, and see where it leads.

http://blog.ioshints.info/2006/01/contact-me.html

Best Regards,

Mike

Nagaraja Thanthry
Cisco Employee
Cisco Employee

Hello,

I think the link that Michael was referring to does not apply to your situation. The link talks about the router interface being dhcp client. In your case, the router is acting as a DHCP server. I had seen similar issues before and they were related to NTP. If you have enabled NTP on the router (DHCP Server) and if for some reason the router is out-of-sync with the NTP server, it might create a situation that you have experienced. When the router realizes that its DHCP database is out of sync, it tries to flush the information in the database and resync forcing all the devices to re-register. Can you check to see if the time changes on your DHCP Server (Router) when the issue happens? Also, check the DHCP database to see if the bindings are still there or they are slowly rebuilding.

Regards,

NT

NT ,

Thanks for the reply, im not running NTP on the router.  The only NTP association is on the voip server which provides Time and Date info to the phones which the clients are connected to, which is making me think that might be knocking them out of sync with the actual router time vs the phone time?

well i went ahead and set up a Windows server as the DHCP server.. still does it.. THIS IS SO ODD!!!!!  I read the whole blog on the way cisco handles dhcp but if i have a windows server running the dhcp , why would it disconnect every single day exactly 24 hours from the time that the lease was established even if i have a 7 day lease set?

Are there any applicable debugs you can run on the int at the time of the behavior?

Edit:

Strike that, I forgot you ran DHCP on another device same issue.

Im running a debug levelthe syslog on the switch which is a SFE2000P , not really getting as much info as i would like out of it.  I cant seem to debug actual ports alone.

Also running debug on the actual phones too now instead of just plain old syslog.   So i will have to see tomorrow i guess when the interface decides to bounce again after 24 hours.  I still think its some type of DHCP extension request problem.. But i can honestly say i've never seen this behavior before.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: