Slaves doesn't handle "expire" time correctly?

[Adam Tkac]

Hi all,

I did some tests with BIND recently and I found problem with zone
expiration:

Master: 9.5.0b1 (fec0::f101:216:3eff:fe38:dec6)
named.conf:
...
zone "lh" IN {
        type master;
        file "lh";
        allow-update { none; };
        allow-transfer { fec0::f101:216:3eff:fe4c:be9d; };
        notify no;
};
...

lh file:

$TTL 1D
@       IN SOA  ns1.lh. rname.invalid. (
                                        6       ; serial
                                        5       ; refresh
                                        5       ; retry
                                        20      ; expire
                                        5 )     ; minimum
@       NS      ns1.lh.
@       NS      ns2.lh.
ns1.lh. IN AAAA fec0::f101:216:3eff:fe38:dec6
ns2.lh. IN AAAA fec0::f101:216:3eff:fe4c:be9d
a.lh.   A       1.1.1.1
---------------------------------------------------------------
Slave: 9.3.3rc2 (I don't see any change related to this to 9.3.5b1)
 - fec0::f101:216:3eff:fe4c:be9d
named.conf:
...
zone "lh" IN {
        type slave;
        file "slaves/lh";
        masters { fec0::f101:216:3eff:fe38:dec6; };
};
...


When I start master and slave zone is transferred to secondary NS as
expected. But network traces & logs shows that secondary doesn't check
master zone every 5 seconds as it should. Also when I stop master and
try query secondary for a.lh. record one minute after master has been
stopped a.lh. record is all along avaliable. Could someone point me
where will be problem?

Regards, Adam

-- 
Adam Tkac, Red Hat, Inc.