Re: Router DHCP Binding Error

awairlines · ‎07-18-2004

Jul 18 22:00:39: DHCPD: dhcpd_age_bindings could not lock semaphore.

I've researched and can't find anything on this error message from a 2651.

Can anyone shed some light on what it means? I have several systems on the remote end that loose their dhcp binding and have to clear ip dhcp binding to get them to release / renew.

Thanks in advance.

Router show ver below.

Cisco Internetwork Operating System Software

IOS (tm) C2600 Software (C2600-IK2S-M), Version 12.1(3)T, RELEASE SOFTWARE (fc1

)

Compiled Wed 19-Jul-00 23:35 by ccai

Image text-base: 0x80008088, data-base: 0x81051CF8

ROM: System Bootstrap, Version 12.1(3r)T2, RELEASE SOFTWARE (fc1)

System image file is "flash:c2600-ik2s-mz.121-3.T"

fmachoka · ‎07-19-2004

Hello,

Below is a document link on troubleshooting DHCP server to determine the source of the error message observed.

http://www.cisco.com/en/US/products/sw/iosswrel/ps1830/products_feature_guide09186a008008743b.html#wp5248

Regards

Fm

allan.thomas · ‎07-20-2004

I came across this error when the bindings were exported to a tftp hostfile, and subsequently read following a restart/reload of the router.

In this particular scenario, an MSFC was providing DHCP, and all bindings were TFTP'd to a remote file, which can be imported following the MSFC being reset.

I found that the particular error was caused by the tftp software, after trying another tftp server the problem was resolved. I don't know whether this applies in your scenario, as I don't know how you have provision dhcp ??

awairlines · ‎07-20-2004

I have 70+ sites each with 2600 routers onsite and each router has a dhcp pool configured to provide addressing locally. The one in question has a scope configured for 100 addresses of which only 70 are being assigned and only 4 are statically being assigned. I have to check the "show ip dhcp conflict" to see if any addresses have been seen as conflicts and then "clear ip dhcp conflicts" because the local users complain that they no longer receive an IP address.

The TFTP server (3COM and Solarwinds) is located at the host site and is only used for storing the startup / running configs and IOS images.

Thanks for any suggestions

allan.thomas · ‎07-20-2004

If the dhcp bindings on each 2600 are not updated to a remote file(s), then when the router reloads it will have no knowledge of any prior dhcp bindings.

Consequently, when a dhcp client reboot/restarts it will request the previous ip address is was allocated. The dhcp server/router will not have any record of this binding and will simply not acknowledge the dhcp request from the client, this is as per the RFC.

The above will not be a problem, unless the router in question reloads. Has the router reloaded ? Have tried configuring a dhcp database agent ?

New clients may be allocated an ipaddress that was previously assigned to another host, because it has no previous bindings, however it is not offered unless there is no icmp reply from the ipaddress it wishes to allocated. Therefore avoiding duplicate ipaddress/conflicts.