Different registered address reported by ns3 and ns17

Pierre

New Member
Hello,

From time to time (once every 6 months), my dynamic address is not correctly registred by Zoneedit and my site becomes unavailable.

Currently, my address is 173.176.182.33, checked on my router and on ZoneEdit control panel.

$ nslookup alterna.tv 8.8.8.8
Server: 8.8.8.8
Address: 8.8.8.8#53

Non-authoritative answer:
Name: alterna.tv
Address: 144.172.184.108

===> Google DNS reports incorrect address!

$ nslookup alterna.tv ns3.zoneedit.com
Server: ns3.zoneedit.com
Address: 128.199.65.59#53

Name: alterna.tv
Address: 173.176.182.33

===> NS3 knows the correct address

$ nslookup alterna.tv ns17.zoneedit.com
Server: ns17.zoneedit.com
Address: 136.0.0.133#53

Name: alterna.tv
Address: 144.172.184.108

===> NS17 still has an obsolete address!!!

Can you resynch the zones of NS3 and NS17?

Thanks
--
Pierre
 

sandy

Administrator
Staff member
at present it looks fine:

dig @ns17.zoneedit.com a alterna.tv

; <<>> DiG 9.6-ESV-R4-P3 <<>> @ns17.zoneedit.com a alterna.tv
; (1 server found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 57806
;; flags: qr aa rd; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0
;; WARNING: recursion requested but not available

;; QUESTION SECTION:
;alterna.tv. IN A

;; ANSWER SECTION:
alterna.tv. 300 IN A 173.176.182.33

;; Query time: 107 msec
;; SERVER: 136.0.0.133#53(136.0.0.133)
;; WHEN: Thu Jul 14 11:46:44 2016
;; MSG SIZE rcvd: 44

Macinsandy - 11:46:44 - ~ > dig @ns3.zoneedit.com a alterna.tv

; <<>> DiG 9.6-ESV-R4-P3 <<>> @ns3.zoneedit.com a alterna.tv
; (1 server found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 46313
;; flags: qr aa rd; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0
;; WARNING: recursion requested but not available

;; QUESTION SECTION:
;alterna.tv. IN A

;; ANSWER SECTION:
alterna.tv. 300 IN A 173.176.182.33


did you make any manual changes?

thanks
sandy
 

sandy

Administrator
Staff member
as well...
dig @8.8.8.8 a alterna.tv

; <<>> DiG 9.6-ESV-R4-P3 <<>> @8.8.8.8 a alterna.tv
; (1 server found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 15258
;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;alterna.tv. IN A

;; ANSWER SECTION:
alterna.tv. 299 IN A 173.176.182.33


sandy
 

Pierre

New Member
Hi Sandy,

To solve that synchronization problem, I had to enter ZoneEdit DynDNS console and do a bogus operation and save the changes. 5 minutes later both DNS zones were identical.

I don't know what creates that problem as my ISP rarely changes the IP assigned to customers. I don't have a log of my IP changes. This problem already happened at least once this year or the previous one. Now everything is fine...

Thanks
--
Pierre
 

sandy

Administrator
Staff member
hmm could be the 0 seconds TTL the dynamic record had. It must be 600 seconds or ten minutes at least. I have updated this for you now.

thanks
sandy
 

gth

New Member
Is there a status page where I can check the DNS propagation for my Zoneedit domain name, on the various zoneedit name servers? An example of what I'm asking is shown at whatsmydns.net

The reason I'm asking is I'm getting the same problem posed earlier in this thread, i.e. -

> server ns1.zoneedit.com
Default Server: ns1.zoneedit.com
Address: 23.27.48.58

> [my domain]
Server: ns1.zoneedit.com
Address: 23.27.48.58

Name: [my domain]
Address: [old IP address]

> server ns3.zoneedit.com
Default Server: ns3.zoneedit.com
Address: 128.199.65.59

> [my domain]
Server: ns3.zoneedit.com
Address: 128.199.65.59

Name: [my domain]
Address: [new IP address]
Checking around the world, everywhere still resolves my domain to the old IP address as listed under ns1, which was from a DDNS API update couple of days ago...
 

sandy

Administrator
Staff member
the domain is on:
@ IN NS ns17.zoneedit.com.
@ IN NS ns3.zoneedit.com
so only those two nameservers would be queried for the domains host records... by chance did you check ns17 as well?

thanks
sandy
 

gth

New Member
At the time of my earlier post, no I didn't check ns17 and whether it had the 'new' or 'old' IP.

Since then, the IP from my ISP has changed again and 13hrs ago (as of writing this post) the API update was successful, however all of the zoneedit nameservers I've checked are still reporting the old address. Are my expectations unrealistic here? How long should ns3 or ns17 take to return the new IP address to an nslookup, after an API update?

Also, not sure why the domain lookup result wouldn't propagate internally at Zoneedit quite quickly, even if I queried any of the other nameservers? How long would that normally take to report an updated IP address?
 

sandy

Administrator
Staff member
I am sorry I cannot find the thread with the domain name.. whats the domain name again?

thanks
sandy
 

gth

New Member
I've sent a PM with the domain name. It's Aug 8 and I can't see any zone edit DNS host that returns IP that was successfully updated on Aug 3 - they're all returning the older IP.
 

sandy

Administrator
Staff member
the domain was reporting properly on its delegated servers when I checked yesterday.

thanks
sandy
 

Pierre

New Member
Hello Sandy,

Same problem again today:

pierre@konrad:~$ nslookup alterna.tv ns3.zoneedit.com
Server: ns3.zoneedit.com
Address: 128.199.65.59#53

Name: alterna.tv
Address: 74.57.162.237

pierre@konrad:~$ nslookup alterna.tv ns17.zoneedit.com
Server: ns17.zoneedit.com
Address: 136.0.0.133#53

Name: alterna.tv
Address: 173.176.182.33

I don't know why the two servers are no more synchronized, but it's causing us problems when we can't access reliably our mail server at home. I'm going to force an update in the admin to refresh the records on the servers.
--
Pierre
 

Pierre

New Member
In case it helps diagnosing the root cause of the problem, it started happening yesterday morning EST (GMT -5). Have you changed something in the servers' configs around that time (for instance applying the latest BIND security patch...)?
--
Pierre
 

sandy

Administrator
Staff member
Ok think we know what going on... the txt record for bug._ is longer then 256 characters, which is the max allowed for a txt record in BIND ...we have a method that modifies these into multiple 256 character records and it should be in place within the next week or so... as a work around to the BIND restriction.

so.. until we have that ready you may need to tweak the SOA to force a reload when the IP changes .. in addition the dyn TTL must be no less then 600 seconds to avoid a TOO SOON error.

thanks and sorry for the inconvenience.

sandy
 

Pierre

New Member
Thanks Sandy,

IP changes don't occur frequently with my ISP and you'll probably have the BIND workaround implemented before next IP change.

I've changed the DynDNS TTL to 600 s in the admin. I hope my IP updater won't reset it.

Thanks for the support.
 

Pierre

New Member
Hi Sandy,

The same problem occurred again today: two different IP addresses registered in two DNS servers. I updated manually the configuration and I noticed that the TTL was again at 0 seconds, while I had changed it to 600 s last time the incident occurred. I've changed it again today in the admin interface. There's no way I can change it dynamically with the IP updates, so something is resetting it.
--
Pierre
 
Top