No subject

10000 a day). Most of them are working as expected. But from time to time
(about 100 times a day) the dhcp-server gets a time-out when sending the
ddns-update. It does happening in all subdomains and in all subnets, not
only in a special zone.

First I thought the problem is on the dhcp-side or somewhere on the network,
but the I traced the networktraffic.

When the dhcp-server gets a timeout he first sent a Update-Packet to the
dns-server (seen with tcpdump) and this packet also arrived as expected on
the dns-server (also seen with tcpdump). But: there is no answer from the
dns nor is there any message regarding this packet in the lofgiles of the
bind. Not in the update.log nor in the error.log nor in the query.log.

The dns.server seemes to ignore the packet.

One idea was, the udp-buffer was full and so the packet was discared by the
system. But when I doubled the UDP-buffersize it changed nothing.

The packets which are ignored look like this (tcpdump):
14:18:48.972335 IP 192.162.3.221.45263 > 192.162.3.100.53:  54647 update
[1a] [2n] [1au] SOA? subzone.mydomain.com. (195)

Again: Most times these packets get the right respone, onle sometimes they
are ignored.

Have anyone any idea what this behaviour causes? (And how to get rid of it?)

By the way: The dns-server is most time 70% idle. So its not a
overloadproblem.

Thanks for any hint,

Tom.

-- 
Lust, ein paar Euro nebenbei zu verdienen? Ohne Kosten, ohne Risiko!
Satte Provisionen für GMX Partner: http://www.gmx.net/de/go/partner