Loop detected! We were referred back to 'XX.XX.XX.XX'
Posted: Wed Jan 25, 2017 3:17 pm
Hello everyone.
I have been using VestaCP for a long time, and I was forced to change the server.
I'm using a new OVH Ubuntu 16.04 Server (64 bits). Yesterday I changed the configuration of my NS in my registrar, and today the mayhem started.
Basically, in my registrar website I have ns1.mysite.com and ns2.mysite.com (being mysite.com my domain) in the section "Personal DNS server", pointing to my IP address. In nameservers, I'm using ns1.mysite.com. and ns2.mysite.com..
Now, in my server, I started purging bind9-host before installing VestaCP, and without purging it, as I read that there was some problems with it.
I do the installation of Vesta CP with the following command:
The panel installs OK, I can access, I go to 'Packages' section and delete all packages but 'default'. I edit 'default' and change the nameservers for ns1.mysite.com and ns2.mysite.com. Then I go to the section 'Web', where I can find that there is a new web configured as panel.mysite.com. I tried both leaving, removing, and going to the 'DNS' section, editing, and choosing child-ns and adding my ns1.mysite.com and ns2.mysite.com and makes no difference.
So I try to add a new website called mysite.com (like panel.mysite.com but just without the word panel). Again, I choose it as child-ns, I'm sure it's using ns1.mysite.com and ns2.mysite.com and eventually I can access to the temporal website (although now is not working and I can't find the reason).
The matter is that when I try to use letsencrypt, no matter if it's visually or through the cli, I always receive the same error: 'Error: DNS problem: SERVFAIL looking up A for mysite.com'. If I use mxtoolbox.com, I receive the error 'Loop detected! We were referred back to 'XX.XX.XX.XX''
The configuration of my DNS is the following:
I have been using VestaCP for a long time, and I was forced to change the server.
I'm using a new OVH Ubuntu 16.04 Server (64 bits). Yesterday I changed the configuration of my NS in my registrar, and today the mayhem started.
Basically, in my registrar website I have ns1.mysite.com and ns2.mysite.com (being mysite.com my domain) in the section "Personal DNS server", pointing to my IP address. In nameservers, I'm using ns1.mysite.com. and ns2.mysite.com..
Now, in my server, I started purging bind9-host before installing VestaCP, and without purging it, as I read that there was some problems with it.
I do the installation of Vesta CP with the following command:
Code: Select all
bash vst-install.sh --force --nginx yes --apache yes --phpfpm no --vsftpd yes --proftpd no --exim yes --dovecot yes --spamassassin yes --clamav yes --named yes --iptables yes --fail2ban yes --mysql yes --postgresql yes --remi yes --quota yes --hostname panel.mysite.com --email [email protected] --password thepassword
So I try to add a new website called mysite.com (like panel.mysite.com but just without the word panel). Again, I choose it as child-ns, I'm sure it's using ns1.mysite.com and ns2.mysite.com and eventually I can access to the temporal website (although now is not working and I can't find the reason).
The matter is that when I try to use letsencrypt, no matter if it's visually or through the cli, I always receive the same error: 'Error: DNS problem: SERVFAIL looking up A for mysite.com'. If I use mxtoolbox.com, I receive the error 'Loop detected! We were referred back to 'XX.XX.XX.XX''
The configuration of my DNS is the following:
- mail._domain...TXT "k=rsa; p=lsakjhdkiashdkahsdkaspaSDASDAUAAS9s89asas88sa8d0a"
_domainkeyTXT "t=y; o=~;"
_dmarcTXT "v=DMARC1; p=none"
@TXT "v=spf1 a mx ip4:XX.XX.XX.XX ?all"
@MX10 mail.mysite.com.
ftpA XX.XX.XX.XX
popA XX.XX.XX.XX
wwwA XX.XX.XX.XX
mailA XX.XX.XX.XX
@A XX.XX.XX.XX
EDIT
DELETE
@NS ns2.mysite.com.
@NS ns1.mysite.com.