Change of internal IP wreaks havoc

While Im a new user, I have been finding answers to my Vmin/Wmin issues for years here and this forum has been a great resource. Thank you.

I am hoping someone can assist with, what I believe, is a simple (feels like a noob) issue that I just cant figure out.

We have a host with single shared IP, that was originally configured for DHCP (unknown why) that lost power. When we brought it back online, it received a different IP from DHCP. No biggie I thought. Change the ip back to what it was and we should be good to go. No changes needed to NAT or external Firewall settings.

We did this and can hit individual servers on this host internally just fine but cannot hit any of them externally. I suspect it is something related to the internal DNS translation but everything looks fine however I am not a BindDNS expert.

Have run ReCheck Config with no issues. We see external traffic hitting the host (i.e no blocks) but cannot load any server on this host externally (Page Cannot Be Loaded). I did try to manually change external IP on one domain to see if that had an impact and it did not. We cleared the iptables as well just in case. No luck.

Any thoughts of what I may be missing here. You assistance would be greatly appreciated.

To get things clear; you can reach the websites fine on the internal network? If that’s the case I would start looking in the router/firewall. While you’re at it; set a static ip-adress in the DHCP configuration for the server.

Thanks for prompt response. Yes I have changed the dhcp config to static with the original ip. I can hit any site on the hoist from internal using our internal (but external to host) DNS which points the domain to the internal ip. It has always been configured this way with no issues. When external traffic hits the host, it just stops and browsers display “domainname took too long to respond”.

Can you ping it externally?


Yes I can with responses.

I had a similar problem, maybe an identical problem, because DHCP overwrote /etc/resolv.conf and removed the nameserver entry. (My host uses DHCP reservations for VPS’s, so the IP is always the same, but it’s dynamically-assigned.)

The fix was to add the nameserver entry to /etc/resolv.conf and make it immutable with chattr +i /etc/resolv.conf. Haven’t had a problem since.

I don’t know if this is your problem; but hey, it’s already broken anyway, so it’s worth a shot.

This was on CentOS 7 and Virtualmin Pro.


Yeah I was thinking similar so I had already looked at resolve.conf. While it does have todays date on it suggesting it was changed in some way, it contains the following which I believe is correct. X’s replace domain for security. 1.200 is legit external to host DNS server.


Had not CHATTR but have now. No change.

You may want to -i that for the time being while troubleshooting.


This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.