Dec 10 14:47:52 server named[17862]: client 97.xxx.xxx.127#50043: view
external: updating zone 'device.example.net/IN': deleting rrset at
'greg.device.example.net' A
Dec 10 14:47:52 server named[17862]: client 97.xxx.xxx.127#50043: view
external: updating zone 'device.example.net/IN': adding an RR at
'greg.device.example.net' A
Which is correct from what I remember the last time I did this.
my zone configuration:
/etc/named.conf
zone "device.example.net" {
type master;
file "/var/named/device.example.net.db";
allow-transfer { any; };
allow-update { key device.example.net; };
};
zone file:
$ORIGIN .
$TTL 3600 ; 1 hour
device.example.net IN SOA ns1.example.net. ns2.example.net. (
2009120805 ; serial
900 ; refresh (15 minutes)
600 ; retry (10 minutes)
86400 ; expire (1 day)
3600 ; minimum (1 hour)
)
NS ns1.example.net.
NS ns2.example.net.
A 205.234.215.112
MX 0 server.example.net.
$ORIGIN device.example.net.
$TTL 60 ; 1 minute
greg A 97.xxx.xxx.127
Running:
BIND 9.3.6-P1-RedHat-9.3.6-4.P1.el5
any suggestions would be welcome. I have run out of ideas and googles.
Secondly, if you do an AXFR of the zone after the Dynamic Update, does
it reflect the change?
Thirdly, on the machine which is originating the ping, how is it set up
to resolve names? Does it only use DNS? Does it only use *itself* for
resolving DNS? Is there some intermediate caching going on (e.g. nscd or
equivalent)? If so, have you waited long enough for the entries to
expire from that intermediate cache?
- Kevin
Hi kevin
Just plain Dynamic Update with "crypto-authenticated" keys
if I do a dig on
root@server [~]# dig @ns1.example.net device.example.net A +tcp
; <<>> DiG 9.3.6-P1-RedHat-9.3.6-4.P1.el5 <<>> @ns1.example.net
device.example.net A +tcp
; (1 server found)
;; global options: printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 44660
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 2, ADDITIONAL: 0
;; QUESTION SECTION:
;device.example.net. IN A
;; ANSWER SECTION:
device.example.net. 3600 IN A 205.xxx.xxx.112
;; AUTHORITY SECTION:
device.example.net. 3600 IN NS ns1.example.net.
device.example.net. 3600 IN NS ns2.example.net.
;; Query time: 1 msec
;; SERVER: 205.234.215.113#53(205.234.215.113)
;; WHEN: Fri Dec 11 03:30:08 2009
;; MSG SIZE rcvd: 85
There should be an A record for a host greg.device.example.net. IN A
97.xxx.xxx.127
Yet if I cat the zone file there is a record
greg A 97.xxx.xxx.127
I'm doing the ping on the dns server that is hosting the
device.example.net zone ..
Thanks for your assistance ..
- Kevin
Or do an "rndc freeze example.net". This will stop dynamic updates to the zone and commit the logfile to the zonefile. Be sure to do an "rndc unfreeze example.net" when you're done to reenable dynamic updates.
"rndc thaw [zone]" is the documented way to resume dynamic updates.
I'd also recommend getting acquainted with "named-journalprint"
(formerly just "journalprint") which will allow you to see the deltas
that have been made to a given zone without taking that zone into
"frozen" state.
AlanC
There is also the -j option of named-checkzone (combined with writing
out a clean version with -D & -o), but unfortunately that only works
if the journal is named using the default "add .jnl to the zone file
name" convention.
Altogether, using AXFR is the thing to get used to using in this
context. (If you disable zone transfers generally, at least allow
them on the loopback interface.) Then start using "masterfile-format
raw", and forget about thinking of zone files are something human
readable ...
--
Chris Thompson
Email: ce...@cam.ac.uk