I administer a web and mail server whose IP was changed some weeks
ago.  Since then I've had intermittent problems, affecting only some
of my users.  They get timeout errors when attempting to follow
certain links on my web pages.

The server is cohosted at my ISP.  The DNS is handled by my domain
registrar, Dreamhost.  (Previously I ran my own BIND server.)  All of
the following looks OK to me:

$ dig @ns1.dreamhost.com -t any pinko.net

; <<>> DiG 9.4.2-P1 <<>> @ns1.dreamhost.com -t any pinko.net
; (1 server found)
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 10719
;; flags: qr aa rd; QUERY: 1, ANSWER: 6, AUTHORITY: 0, ADDITIONAL: 4
;; WARNING: recursion requested but not available

;; QUESTION SECTION:
;pinko.net.                     IN      ANY

;; ANSWER SECTION:
pinko.net.              14400   IN      A       208.79.2.3
pinko.net.              14400   IN      NS      ns2.dreamhost.com.
pinko.net.              14400   IN      NS      ns1.dreamhost.com.
pinko.net.              14400   IN      SOA     ns1.dreamhost.com. 
hostmaster.dreamhost.com.
2008082600 20690 1800 1814400 14400
pinko.net.              14400   IN      MX      10 debs.pinko.net.
pinko.net.              14400   IN      NS      ns3.dreamhost.com.

;; ADDITIONAL SECTION:
ns2.dreamhost.com.      14400   IN      A       208.96.10.221
ns3.dreamhost.com.      14400   IN      A       66.33.216.216
debs.pinko.net.         14400   IN      A       208.79.2.3
ns1.dreamhost.com.      14400   IN      A       66.33.206.206

;; Query time: 65 msec
;; SERVER: 66.33.206.206#53(66.33.206.206)
;; WHEN: Tue Aug 26 19:10:19 2008
;; MSG SIZE  rcvd: 259


Now let's query the TLD name servers for debs.pinko.net.  Note that
the A record returned for debs.pinko.net is different; that is the old
IP address.


$ dig @a.gtld-servers.net debs.pinko.net

; <<>> DiG 9.4.2-P1 <<>> @a.gtld-servers.net debs.pinko.net
; (1 server found)
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 26253
;; flags: qr rd; QUERY: 1, ANSWER: 1, AUTHORITY: 3, ADDITIONAL: 3
;; WARNING: recursion requested but not available

;; QUESTION SECTION:
;debs.pinko.net.                        IN      A

;; ANSWER SECTION:
debs.pinko.net.         172800  IN      A       204.96.181.68

;; AUTHORITY SECTION:
pinko.net.              172800  IN      NS      ns1.dreamhost.com.
pinko.net.              172800  IN      NS      ns2.dreamhost.com.
pinko.net.              172800  IN      NS      ns3.dreamhost.com.

;; ADDITIONAL SECTION:
ns1.dreamhost.com.      172800  IN      A       66.33.206.206
ns2.dreamhost.com.      172800  IN      A       208.96.10.221
ns3.dreamhost.com.      172800  IN      A       66.33.216.216

;; Query time: 74 msec
;; SERVER: 192.5.6.30#53(192.5.6.30)
;; WHEN: Tue Aug 26 19:13:02 2008
;; MSG SIZE  rcvd: 163


Albitz & Liu, DNS and BIND, says this indicates a delegation problem,
if I understand them correctly.  So does this:

$ dig @a.gtld-servers.net -t ns pinko.net

; <<>> DiG 9.4.2-P1 <<>> @a.gtld-servers.net -t ns pinko.net
; (1 server found)
;; global options:  printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 32942
;; flags: qr rd; QUERY: 1, ANSWER: 3, AUTHORITY: 0, ADDITIONAL: 3
;; WARNING: recursion requested but not available

;; QUESTION SECTION:
;pinko.net.                     IN      NS

;; ANSWER SECTION:
pinko.net.              172800  IN      NS      ns1.dreamhost.com.
pinko.net.              172800  IN      NS      ns2.dreamhost.com.
pinko.net.              172800  IN      NS      ns3.dreamhost.com.

;; ADDITIONAL SECTION:
ns1.dreamhost.com.      172800  IN      A       66.33.206.206
ns2.dreamhost.com.      172800  IN      A       208.96.10.221
ns3.dreamhost.com.      172800  IN      A       66.33.216.216

;; Query time: 79 msec
;; SERVER: 192.5.6.30#53(192.5.6.30)
;; WHEN: Tue Aug 26 19:17:04 2008
;; MSG SIZE  rcvd: 142


I understand the parent shouldn't be handing out NS records in the
answer section, rather, they should appear in the authority section.


Verisign says "talk to Dreamhost."  Dreamhost says "our servers are
functioning properly."  If this is indeed bad delegation, what's the
right way to solve the problem?  Or is my diagnosis wrong?


Reply via email to