domains not working after OS upgrade

V

verruckt

Verified User
Joined
Mar 8, 2004
Messages
52
I have a server that has OS corruption, and we determined the correct thing to do would be to upgrade the OS. We did that, and everything was ok with the OS. All the user data was left intact. The server is now up but, DA did not automatically start, I had to start it manually. None of the sites are up. You can get to them via IP, but not by name. Do I need to rerun the DA setup script? Or is there a simple configuration change that I'm overlooking? Thanks for any help.
 
dannygoh said:
What are you upgrading from? More details is need in order to help.
Click to expand...

Sorry. This server had centos3.5, and upgraded to 3.8. Nothing was changed with regards to the home directories, or directadmin. Is it possible that I need to run the DA install script? DA still shows everything as it should be, but I can only get to the sites or email via IP/user.
 
Last edited:
How did you upgrade CentOS? Did you allow the CentOS upgrade to install apache? If so you need to remove it, and run the customapache scripts (search these forums).

Jeff
 
I have an issue where I did YUM updates and it appearantly updated BIND to version 9.2.4. After this update, no domain names on my server were able to be resolved. The named service *IS* working and I did some investigation and found this thread and even followed the instructions regarding the above link of moving the *.db files to /var/named. I did that and restarted named, to no avail.

All of the domain info *IS* there, but it's like it's unable to read any of it.

Help!!!
 
verruckt said:
We looked into that, but there is no /etc/named/chroot directory. Perhaps it could be something else?

Also, in named.conf all of the domains and IP's are in there, along with the associated db files.

I'm stumped. :(
Click to expand...
It would actually be in the /var/named/chroot directory and not /etc/named/chroot. Also can type in
Code: 
ps ax | grep named
and see if it outputs something like
Code: 
/usr/sbin/named -u named -t /var/named/chroot
if you have the -t flag on it means its chrooting it somewhere and you need to make sure the .db files are there. Also as Jeff said use vi to check your /etc/httpd/conf/httpd.conf and make sure all the includes are at the bottom for each of the different user's and their domains.
 
I found the issue... it overwrote the /etc/named.conf file. I had to restore one from backups because all of the linked client zone files was missing. I did that restarted the named service and it works great.
 
to keep it from happening again be sure that yum won't update the bind-chroot package.

Jeff
 
You must log in or register to reply here.
Back
Top