Odd AXFR issue

Buddy Ellis Buddy at southwest.cc.nc.us
Wed May 18 13:52:05 UTC 2005


I'm having some strange issues with bind 9.3.1.

My master is running on SuSE 9.2, firewall is currently turned off, its IP =
address is 10.210.1.53.
My secondary is running on solaris 9, IP address is 10.210.1.18

Both are running 9.3.1.

On the master I have a zone entry:

zone "mydom.com" {
        type master;
        file "/conf/mydom.com.hosts";
        notify yes;
        };


On the slave I have an entry:

zone "mydom.com" {
        type slave;
        file "/conf/mydom.com.hosts";
        masters { 10.210.1.53; };
        };


Whenever I do a zone update, and rndc reload, the master sends out a =
notify to .18 as expected:
18-May-2005 09:19:51.567 notify: info: client 10.210.1.18#53: received =
notify for zone 'mydom.com'

However, the secondary then refuses to accept the notify because it is a =
'non-master',and it seems to be identifying the machine doing the notify =
as its own IP address, again, the master doing the notify is .53. and the =
slave is .18:

18-May-2005 09:19:51.567 general: info: zone mydom.com/IN: refused notify =
from non-master: 10.210.1.18#53

If I add an entry for .18 in the secondary server as a 'master' it accepts =
the AXFR:

18-May-2005 09:19:52.063 xfer-in: info: transfer of 'mydom.com/IN' from =
10.210.1.53#53: connected using 10.210.1.18#53261
18-May-2005 09:19:52.097 general: info: zone mydom.com/IN: transferred =
serial 2005051805

Now if I restart .18 and 'force' a reload, it transfers fine:

18-May-2005 09:39:25.841 general: info: zone mydom.com/IN: Transfer =
started.
18-May-2005 09:39:25.843 xfer-in: info: transfer of 'mydom.com/IN' from =
10.210.1.53#53: connected using 10.210.1.18#53267
18-May-2005 09:39:25.875 general: info: zone mydom.com/IN: transferred =
serial 2005051805
18-May-2005 09:39:25.876 xfer-in: info: transfer of 'mydom.com/IN' from =
10.210.1.53#53: end of transfer

I've been banging my head up against the wall for a few days trying to =
figure out what the heck the problem is here.  It almost seems like I have =
some weird IP mangling going on or something, where the master IP address =
is getting munged to be .18, but there is nothing between the master and =
slave that could be doing that, as they are just attached to the same =
switch, on the same net.

Any insight into what could be going on here would be greatly appreciated =
in advance. =20

-b







More information about the bind-users mailing list