I have spent weeks reading and trying things and cannot resolve this. I always try to sort it myself before asking for help. If I ask for help it is because I am stuck. I am currently reading about how to flush the caches to see if that helps. I have also searched the entire server for references of that IP address and it reports that it does not exist. perhaps the grep command I used was wrong? I tried:
Also, when I run it I get a long list of “permission Denied”, “Invalid argument” and “Input/output error” listings so it is not searching those files. how do I make the search get inside those file (I am logged in as root)?
Cleared bind cache and reloaded. Inserted my DNS Servers to network configuration and applied the configuration. It returns with the single incorrect server!!!
I have now checked and read through all startup services and tried to understand systemd but I cannot find any reference to the server establishing 213.186.33.99 as my dns server. I cannot move forward without some clear help.
Well I have managed a workaround that prevents resolve.conf being overwritten by using
chatter +i resolv.conf
but this is a plaster on a boil and I want to lance the boil so I would still appreciate assistance on it.
I am now going to try to discover why bind is not showing my zone. Any guidance please on where to start?
As with resolve.conf above, I have managed a solution but it does not explain why this is happening.
To overcome the problem I;
Opened the zone file in /var/lib/bind and copied all the entries to an editor
Opened the Bind config /etc/bind/named.config.local and copied all the entries to the same editor
Deleted the zone file in /var/lib/bind
Deleted the entries in /etc/bind/named.config.local (I only have one virtual server so this emptied the file).
This finally allowed me to add a master zone for xorex.rocks
I then edited the zone file in /var/lib/bind and deleted the entire content before pasting the previously automatically generated entries (plus my single manual addition).
Saved and closed the file and suddenly xorex.rocks was listed as a master zone and is working and accessible. It immediately showed me why xerxes.xorex.rock was not working as I have not added the final period so the entry was xerxes.xorex.rocks.xorex.rocks. I have amended it to be correct and will wait to see if it now resolves.
I then created a new virtual server test.xorex.rocks and it immediately became visible in bind.
The original (non-showing) entry generated in /etc/bind/named.config.local was as follows:
zone “xorex.rocks” {
type master;
file “/var/lib/bind/xorex.rocks.hosts”;
allow-transfer
127.0.0.1;
localnets;
213.251.188.141;
2001:41d0:1:4a8d::1;
};
};
The replacement was simpler:
zone “xorex.rocks” {
type master;
file “/var/lib/bind/xorex.rocks.hosts”;
};
Yet the new virtual server is visible with the entry:
zone “test.xorex.rocks” {
type master;
file “/var/lib/bind/test.xorex.rocks.hosts”;
allow-transfer {
127.0.0.1;
localnets;
213.251.188.141;
2001:41d0:1:4a8d::1;
};
};
and if I then re-insert the original entry and restart bind it still shows the zone so this is not what is causing the original problem.
Joe, I have read lots of posts on google with this problem and all are using a workaround. I believe it is either a virtualmin problem or a compatibility issue with Debian 10. Is this something you could look into please?
If an IP gets written within the BIND service, its most likely a setting from your hoster and from virtualmin. I think its more likely to be the first.
bind9 1:9.11.5.P4+dfsg-5.1 amd64 Internet Domain Name Server
Sitting on Debian 10
I contacted the hoster and they just do not want to know! Have not even responded to my ticket despite bumping it several times. If I have disabled the cloud initiation services, and flushed the cache for bind, how can it be rewriting the resolv.conf file? the behaviour is quite unusual and I cannot think why it is happening.
If your hoster is pre-setting that in his system, then you cant change it. I am currently not sure if its really that or something you did.
You can try to install the latest BIND9 version from the backports repo.
It is currently updated through backport. I now have it working and am going to leave it for now. I was just hoping that @Joe would look at the post and see if it is a virtualmin issue on Debian 10.
I have now solved both the initial problems. Bind is working as it should, my workaround has reset the dns settings and today I successfully organised the SSL certs after rebuilding the virtual server once again! Not all sub entries are yet on the certificate as both mail and webmail currently make it fail but I will wait once more and post something new if it is still an issue. I have copied the cert I have generated to the various modules so hopefully all should now work.
The next task is to set up the mail servers. Such fun! (You need to be a fan of Miranda, a UK comedienne, to appreciate that).