[Dnsmasq-discuss] Dhcp discover client source ip not 0.0.0.0

Simon Kelley simon at thekelleys.org.uk
Tue May 22 12:00:36 BST 2012


On 22/05/12 11:51, Sen Haerens wrote:
> Greetings,
> 
> I have a Netgear switch that is unable to acquire a DHCP lease. The
> DHCP discover request is not replied according to dnsmasq & tcpdump
> log. I guess the UDP src ip is a violation of the RFC.
> 
> 11:26:06.106723 IP (tos 0x0, ttl 64, id 380, offset 0, flags [none],
> proto UDP (17), length 281)
>    192.168.0.239.68 > 255.255.255.255.67: [udp sum ok] BOOTP/DHCP,
> Request from 74:44:01:78:e1:ed, length 253, xid 0x8274d270, Flags
> [Broadcast] (0x8000)
>          Client-Ethernet-Address 74:44:01:78:e1:ed
>          Vendor-rfc1048 Extensions
>            Magic Cookie 0x63825363
>            DHCP-Message Option 53, length 1: Discover
>            MSZ Option 57, length 2: 576
>            Parameter-Request Option 55, length 3:
>              Subnet-Mask, Default-Gateway, Domain-Name-Server
> 
> If I add a route for this IP range on the server the problem is fixed.
> sudo route add -net 192.168.0.0 netmask 255.255.255.0 dev vlan20
> 
> Shouldn't the broadcast flag in the DHCP request take care of this situation?
> Thank you for your assistance.
> 

Do you know where the faulty request is getting lost? Does dnsmasq log
receiving the DHCPDISCOVER?

If you run strace, does dnsmasq get the discover packet and attempt to
send a reply?

My fear is that the request is being dumped by the kernel before it even
gets to dnsmasq, if that's the case the broadcast flag won't help at all.



Cheers,

Simon.






More information about the Dnsmasq-discuss mailing list