I am attempting to cluster a server on webmin and bind and running into the old “is not running Webmin version 1.202 or above.”
I have gone through a dozen different articles here and on the web and all of them seem to point to a problem in the firewall. Ive opened ports 10000-10014 on both TCP and UDP. Server A is running Debian 9, Server B using FirewallD is running Ubuntu 18.04LTS using Linux Firewall.
I am able to connect it in the webmin cluster page and clicking on its button loads its control panel in the screen, however, I cannot for the life of me get it to connect to auto-slave BIND from Server A.
The only ports you need to open are 53TCP/UDP for DNS, and your Webmin port and Webmin port + 1 (ie. 53, 10000, 10001). If you’re like me, you’ll restrict the IPs that can access these ports to the other servers in your cluster, and your own personal/work network.
Don’t use the Webmin Cluster stuff for this. Here’s how to do it.
On the SLAVE server/s, create a new Webmin user. Give it only access to BIND. Use a ridiculously unguessable username and password. Other than this connection, you’ll never use it (that means, don’t use “root”!).
On the MASTER server, go to Webmin -> Webmin Servers Index and click “Register a new server”.
Fill in the details using the new user/s you just created. You should end up with an icon that if you click it, it logs you into your slave server and you can access BIND.
Once that works, on the MASTER simply go to Webmin -> Servers -> BIND DNS Server -> Cluster Slave Servers. Follow your nose from here and you should be good to go.
is the slave managed in webmin by main server under >webmin>webmin servers index
On main server, go to > servers > Bind Dns server > cluster slave servers. Should give you list of servers to add as slaves,pick server. Change “Create secondary on slave when creating locally” to yes, and “Create all existing master zones on slave?” to yes. Click add now.
On main server still. > servers > bind dns server > zone defaults. In bottom left, “Also notify slaves…” click “listed” check box and put in slave IP(s). Also be sure “Notify slaves of changes?” is set to yes. This tells main to notify slaves of dns changes. So main is now setup.
On the slave(s), > servers > bind dns server > zone defaults. Set “Allow transfers from…” to the IP of the main dns server. Now the “zone defaults” are setfor bind itself. Restart bind on slave, then restart bind on main. Should see new zones on the slave now.
Now have to clear “zone defaults” for individual zones(domains). On the slave(s), > servers > bind dns server > domain name> zone defaults. The 2 “default” check boxes should be checked and all four text boxes should be blank. I believe “127.0.01, local” defaults to one of those boxes, which wont work. Restart the slave Dns and cluster should now work. (NOTE) You’ll have to manually be sure each domain zone doesn’t have the 127.0.0.1,local host each time a new zone is created.
@kilmarac did you resolve this problem? I see lots different answers in this post, and on the internet. On my server I have port 10000 and 53 UDP and TCP open, this all seems good but I still get a long timeout and then that message. This post suggested 10014 but also adding UDP/TCP 10014 did not work:
Yes, my bad. I have updated my original post to include the three ports. I had obviously forgotten about the Webmin+1 port when I originally wrote my post (I don’t use 10000 & 10001 on my own servers). Thanks for the reminder.