[Dnsmasq-discuss] DNS request timeout problem on dial-in account with dynamic ip

Jean Wolter jw5-dns at os.inf.tu-dresden.de
Thu Feb 9 11:56:24 GMT 2006


Simon Kelley <simon at thekelleys.org.uk> writes:

> It's a long time since I played with DoD, but I seem to remember that
> the magic to fix this is
>
> echo 1 >/proc/sys/net/ip_dynaddr

This applies to TCP connections only. See
http://www.linuxdocs.org/HOWTOs/Adv-Routing-HOWTO-12.html

/proc/sys/net/ipv4/ip_dynaddr

    You need to set this if you use dial-on-demand with a dynamic
    interface address. Once your demand interface comes up, any local
    TCP sockets which haven't seen replies will be rebound to have the
    right address. This solves the problem that the connection that
    brings up your interface itself does not work, but the second try
    does.

I did check this in the 2.4 kernel sources and the only references to
this variable are in the handling of TCP packets. And since tcp
handles lost packets by itself ... Unfortunately we have UDP here.

> Jean Wolter wrote:
>> Is there any chance to automatically re-send all un-replied DNS requests
>> after a dial-in happened? I tried to browse through the source but I
>> did not find the right spot to add such a resend operation.
>
> It's impossible to make dnsmasq do this: once it has forwarded a DNS
> query, it no longer has a copy of it, so it can't resend it unless the
> original source of the query does a retry.

If I would like to add an online test before a query is forwarded
and would like to trigger a dial-in explicitly - would forward_query
be a good point to insert such a test or should this be done somewhere
else?

regards,
Jean



More information about the Dnsmasq-discuss mailing list