[AusNOG] Strange DNS issue

[Sean K. Finn]

Gary,

I'll pose some obvious question, do the nameservers have appropriate glue records in the root, (or CC root) and are the domain's NS Records matching the nameservers they are actually delegated to?

You can check TLD nameserver glue here:

http://www.internic.net/whois.html

Or, if the nameservers are .au nameservers, here

http://whois.ausregistry.com.au/whois/whois_local.jsp?

With the query string being "HOST ns1.domain.com.au"

Glue (HOST RECORD) response example for  "HOST NS1.net.au"

Whois response for HOST ns1.net.au:
Host ID

H0041281-AR

Host Name

ns1.net.au

IP Address

202.125.32.4


Sometimes when the glue isn't correct, and when the NS records don't match where the domain name is delegated to, 'strange' things happen, where the domain works from some places, but not others.

Sean.
From: ausnog-bounces at lists.ausnog.net [mailto:ausnog-bounces at lists.ausnog.net] On Behalf Of CAS Netlink Support
Sent: Sunday, October 28, 2012 12:41 PM
To: ausnog at lists.ausnog.net
Subject: Re: [AusNOG] Strange DNS issue

Thanks Heinz. I lowered the refresh time yesterday and I've just seriously increased the serial. It has become a bit curiouser though. Telstra's lon-resolver.telstra.net appears to be returning the correct record when queried from a USA server:

# dig @203.50.2.71 xxxxx.net any

; <<>> DiG 9.3.6-P1-RedHat-9.3.6-20.P1.el5_8.5 <<>> @203.50.2.71 xxxxx.net any
; (1 server found)
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 19922
;; flags: qr rd ra; QUERY: 1, ANSWER: 6, AUTHORITY: 2, ADDITIONAL: 4

However, when I query the same server from my client's firewall which is connected to Bigpond ADSL, it fails:

t# dig @203.50.2.71 xxxxx.net any

; <<>> DiG 9.5.1-P2 <<>> @203.50.2.71 xxxxx.net any
; (1 server found)
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: SERVFAIL, id: 40361
;; flags: qr rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0

I've never seen this sort of weirdness before. DNS seemed so much easier when Robert Elz was running the show :).

Gary


On 28/10/2012 1:12 PM, Heinz N wrote:
Am I right in thinking it is probably that the domain had a long expire time? The refresh was set to 86400, but it seems like it is being ignored or it doesn't figure in how long DNS servers will wait before refreshing the domain.

A few years ago I hade a similar problem when redelegating domains where telstra was the secondary. I had a too long refresh time and it seemed that their DNS only updated according to the longest number: refresh or retry (this is just my opinion). Everyone else's DNSs queried my authorative host and got the redelegations but the secondary (telstra) didn't for quite some time. This problem was my fault and now I have everything set at 1hr (except expire which is set much longer).

I resorted to adding new host A records into those domains and doing a dig @nsX.telstra.XXXX on those host.domain in order to force their DNS to re-query the zone records. This didn't trigger a zone transfer unfortunately, but the new hosts did then appear. I used them until the full zone transfer finally happened. (This is a bit difficult if your "www" A record is the one not updating).

PS. Don't forget to update your Zone record serial number. You might even try increasing it in order to try to trigger a full refresh.

Regards,
Heinz N.

_______________________________________________
AusNOG mailing list
AusNOG at lists.ausnog.net<mailto:AusNOG at lists.ausnog.net>
http://lists.ausnog.net/mailman/listinfo/ausnog

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ausnog.net/pipermail/ausnog/attachments/20121029/0d587fb7/attachment.html>