DNS zones not reloading...

[DCH]

Is anyone else having this problem...

When copying multiple accounts via WHM between two DNS clustered servers....

the IP in the zone gets updated properly. However, named doesn't seem to be reloading the updated zone.

I have to edit the zone - and save without changes. At that point, the IP change takes effect.

Anyone have this problem and know how to fix it?

Thanks!

[mjost]

I have the same issue since I updated to latest current.

When calling the reload from perl with cPScript::DnsUtils::askdnsadmin("RELOADBIND",0,"domain.com");

i get :

Can't use string ("CentOS release 4.3 (Final)
") as an ARRAY ref while "strict refs" in use at /usr/local/cpanel/Cpanel/DNSLib.pm line 831.
Bind reloading on server2 using rndc zone: [domain.com]

it's not reloading the local zone, but it's working on the other server of the cluster (which isn't running the latest current)

Anyone has a solution ?

[jclarke]

I'm also having the same issue, I have a ticket in with cPanel, so hopefully they will have a fix for this soon.

[jackie46]

Yep, that happens here on clustered Freebsd server too.

[jackal]

Having same issue on 5 servers we have tried so far /scripts/fixndc we do not have servers clustered but all have centos 3.7 to 4.3 on them

[mjost]

the issue seems to be only on CentOS : http://bugzilla.cpanel.net/show_bug.cgi?id=4155

[jackal]

Edge 100 still has the issue

root@ [~]# /scripts/fixndc
Can't use string ("CentOS release 3.7 (Final)
") as an ARRAY ref while "strict refs" in use at /scripts/cPScript/DNSLib.pm line 831.

[harpon]

I have the same issue. If anyone has a solution, please let us all know.

Clients are adding domains daily and I have to manually reload bind.

host 1:

WHM 10.8.0 cPanel 10.8.2-C101
CentOS 4.2 i686 - WHM X v3.1.0

Host 2:

WHM 10.8.0 cPanel 10.8.2-C101
CentOS 3.7 i686 - WHM X v3.1.0

Thanks.

[CapTech]

Following the directions I found on a bug ticket, I got our server working properly by commenting out the following line from the /scripts/cPScript/DNSLib.pm file.

           return ($loc,(split(/\//,$loc))[$#${_}]) if wantarray;

To read:

#           return ($loc,(split(/\//,$loc))[$#${_}]) if wantarray;

Then I reloaded bind from WHM. Worked like a charm, and it fixed a major problem we had with a new domain account not resolving.

Regards,

CapTech

[patryk]

Same problem here with lastest edge.. ;/

[chirpy]

Yup, a search in bugzilla will find you the entry I logged about this issue and the workaround I posted that CapTech has quoted above.

[harpon]

I applied the workaround and restarted bind in WHM. Now, when modifying records, it doesnt say that it's restarting bind on the clustered servers. How do i know it's helping at all?

Thanks.

[patryk]

chirpy: Yes, after editing both files mentioned in bugzilla it works .

[harpon]

Thanks guys. Editing both files sorts it out, as Jonathan mentioned.

[Sergiu Tot]

I get the same error when I transfer an account from one web server (CentOS WHM 11.15.0 cPanel 11.17.0-R19434) to another one (FreeBSD WHM 11.15.0 cPanel 11.17.0-R19434).

Bind reloading on ro02 using rndc zone: [domain.ro]
Changed all instances of [OLD_IP] -> [NEW_IP] in domain.ro.db
Bind reloading on ro02 using rndc zone: [addondomain1]
Changed all instances of [OLD_IP] -> [NEW_IP] in addondomain1.db
Bind reloading on ro02 using rndc zone: [addondomain2]
Changed all instances of [OLD_IP] -> [NEW_IP] in addondomain2.db
Bind reloading on ro02 using rndc zone: [domain.ro]
Changed all instances of [OLD_IP] -> [NEW_IP] in domain.ro.db
Bind reloading on ro02 using rndc zone: [addondomain1]
Changed all instances of [OLD_IP] -> [NEW_IP] in addondomain1.db
Bind reloading on ro02 using rndc zone: [addondomain2]
Changed all instances of [OLD_IP] -> [NEW_IP] in addondomain2.db

I tried to modify the /scripts/cPScript/DNSLib.pm file on both CentOS and FreeBSD machine but the result is the same.