Package: bind9
Version: 1:9.8.4.dfsg.P1-6+nmu2
If slave zone expires and then later is available for transfer,
zone gets transferred successfully (and is served correctly), but the
slave zone
file is not updated. Should the bind9 be restarted, old (expired) zone
is loaded from file.
Steps to reproduce:
on master,
named.conf
--
zone "test.example" {
type master;
allow-transfer { any; };
also-notify { <slave>; };
file "/tmp/test/db.test.example";
};
db.test.example:
--
$TTL 900
@ IN SOA ns1 admin.test.example. (
2013060000 ; serial
60 ; refresh
60 ; retry
300 ; expire
900 ; minimum
)
@ IN NS ns1
ns1 IN A 127.0.0.1
@ IN TXT "test1"
--
on slave,
named.conf
zone "test.example" {
type slave;
masters { <master>; };
allow-query {any;};
file "/tmp/test/db.test.example";
};
zone is transferred, /tmp/test/db.test.example created (serial 2013060000).
Don't allow slave to transfer from master:
allow-transfer { 127.0.0.1; };
Await zone expiration on slave:
named[15917]: zone test.example/IN: expired
Update master db.test.example to serial 2013060002, allow slave to
transfer from master,
slave server transfers zone:
Transfer completed: 1 messages, 5 records, 160 bytes, 0.001 secs (160000
bytes/sec)
Slave answers queries with new zone data, but file
/tmp/test/db.test.example at slave server
still contains old zone (serial 2013060000).
Stopping/starting slave bind9 results in old file being loaded, slave
serves test.example with serial 2013060000:
named[28541]: zone test.example/IN: loaded serial 2013060000
Running "rndc retransfer test.example" transfers zone and also updates
zone file.
Bug seems to also affect 1:9.7.3.dfsg-1~squeeze10.
Best Regards,
Tomas
--
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]
