![tpnsolutions's picture tpnsolutions's picture](https://archive.virtualmin.com/sites/default/files/styles/thumbnail/public/pictures/picture-7592-1487694249.png?itok=0emRun-C)
Hi,
I have been testing a new script I wrote whose job is to run the "modify-slavedns" api command on a few DNS servers.
While the command executes without any errors, there is one problem.
Once the new "master" IP address has been set on the slave machines, the zone is not attempted to be transferred. In order to get a transfer to complete I need to goto the "Primary DNS" server and update the serial, then apply the settings which causes an instant transfer. This extra task is not desired.
I've confirmed that on the "Server Configuration > Configure Slave DNS Zone" page that the "Master server IP addresses" is being updated upon the completion of the command.
However as noted, the corresponding "Records transferred from master" is not reflecting the updated zone on the new server.
As a result, I have issued a temporary work around:
Issue "delete-domain" command.
Issue "create-domain" command.
This effectively drops the domain, then re-creates it causing the zone from the new master to transfer instantly.
Please let me know if there is something misconfigured on my end causing this behaviour or kindly implement a fix for this asap.
-Peter
Comments
Submitted by JamieCameron on Thu, 08/11/2011 - 01:36 Comment #1
That is unusual, as Virtualmin should restart BIND on the slave after changing the master IPs, which should trigger a new zone transfer.
If you perform a BIND restart with a command like
/etc/init.d/named restart
after changing the master IP, does that force a transfer?Jamie,
Negative, I did a series of tests by issuing the command via the command line to see the effect.
I ran the command which says it was restarting bind, but no transfer was starting...
I ran
/etc/init.d/named restart
manually after running the API command and it was reverting back to the older master ip record.-Peter
Submitted by JamieCameron on Thu, 08/11/2011 - 14:32 Comment #3
Ok, I will update the slave DNS plugin to wipe that file and do a BIND restart to force the transfer .. that appears to fix the issue in my tests.
Jamie,
Sounds good :-)
-Peter
Submitted by Issues on Thu, 08/25/2011 - 16:24 Comment #5
Automatically closed -- issue fixed for 2 weeks with no activity.