Delays getting an address after a DHCPNAK

Tom Harris thom.j.harris at gmail.com
Fri Jul 19 23:47:56 UTC 2013


On Fri, Jul 19, 2013 at 3:23 PM, Doug Barton <dougb at dougbarton.us> wrote:

> On 07/19/2013 03:20 PM, Tom Harris wrote:
>
>> I have what might be a somewhat uncommon situation (cloning from a VM
>> that has a valid lease), which I've been unable to eliminate a 10 second
>> delay on.
>>
>> The scenario is:
>>
>> - I'm cloning linux containers/VMs in a very transient way.  They come
>> up and down frequently.
>> - The new VM is cloned from a running VM which has a valid lease.
>> - The new VM tries to renew the lease on the IP from the original VM,
>> and gets NAK'd.
>>   - {Delay 10 seconds}
>> - The new VM does a DHCPDISCOVER and quickly gets a valid IP.
>>
>> I modified the reboot timeout, and the initial-interval, but they didn't
>> effect the 10 second delay.   Is there another setting that would effect
>> this?  Or any other way to eliminate the delay?
>>
>
> Have you monitored the client behavior in this state to see if it is the
> client which is waiting 10 seconds after the NAK?
>
>
Yes, I used dhcpdump to monitor the requests, and I see the Client request,
followed immediately by the server NAK.    Then, it's silence for ~10s, and
the client then sends a discover and the server responds appropriately.

So, it does seem to be only on the client side that the wait happens.


> ______________________________**_________________
> dhcp-users mailing list
> dhcp-users at lists.isc.org
> https://lists.isc.org/mailman/**listinfo/dhcp-users<https://lists.isc.org/mailman/listinfo/dhcp-users>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.isc.org/pipermail/dhcp-users/attachments/20130719/aacc978a/attachment.html>


More information about the dhcp-users mailing list