name servers keep failing

InterNext

New member
Joined
Oct 18, 2003
Messages
4
name servers keep failing - 3 failurs in 9 hours

I recently had my datacenter install directadmin on my former cp server. fresh install over rh9.

3 times today I went to the server hoping to start moving accounts over, only to find the nameservers have failed. I can only access the box through ip numbers. no failure notices emailed to me.

1pm, I called my datacenter.. they restarted something and I was back up immediately, able to reach the box via the domain name.

7pm, same problems.. here is what DC told me : "Matt Terry : as i said, I created a symlink that should correct it.
Matt Terry : ln -s /usr/sbin/rndc /usr/sbin/ndc "

Well that didnt work for long. I caught it down again at 10pm, and it is currently still down (trouble ticket in). I moved from cpanel because I heard how more stable and less buggy DA is. Not being able to see my sites 3 times within 9 hours is something I never experienced before with CP. Please help.

Thanks
Eric

www.hostingplans.com
 
Last edited:
First try to restart BIND

# service named restart

also check your log file - /var/log/messages

Take a look @ your named.conf for anything that appears incorrect also.

Chris
 
Restarting BIND... that is what they did the first time (thanks for jogging my memory). I will look into the rest. Thanks.
 
update

My datacenter had two techs look into this problem for me and I was told " the name server configs are fine, and the nameserver runs fine until DA messes with it. It also restarts fine when done manually from the DA web interface."

"Posting a message in the DA forums was a good place to start. If you'd like to add more info to your forum question, this is from the log file:"

Jan 26 00:02:19 quicksilver su(pam_unix)[3984]: session opened for user
diradmin by root(uid=0)
Jan 26 00:02:19 quicksilver su(pam_unix)[3984]: session closed for user
diradmin
Jan 26 00:02:25 quicksilver su(pam_unix)[4008]: session opened for user
diradmin by root(uid=0)
Jan 26 00:02:29 quicksilver named: named shutdown failed

"We are linux experts, but we did not write the DA software, and cannot possibly know why it keeps killing the nameservers."

----------------

The above is what I am up against. Is this a known issue or something that can be fixed? The datacenter was attempting to support DirectAdmin at my request. They normally only offer 'that other control panel' but was going to start offering DA as well if everything went well with mine and on their test server.

Can anyone help? Otherwise its back to CP (and I dont want to do that)
 
Re: update

InterNext said:
My datacenter had two techs look into this problem for me and I was told " the name server configs are fine, and the nameserver runs fine until DA messes with it. It also restarts fine when done manually from the DA web interface."

"Posting a message in the DA forums was a good place to start. If you'd like to add more info to your forum question, this is from the log file:"

Jan 26 00:02:19 quicksilver su(pam_unix)[3984]: session opened for user
diradmin by root(uid=0)
Jan 26 00:02:19 quicksilver su(pam_unix)[3984]: session closed for user
diradmin
Jan 26 00:02:25 quicksilver su(pam_unix)[4008]: session opened for user
diradmin by root(uid=0)
Jan 26 00:02:29 quicksilver named: named shutdown failed

"We are linux experts, but we did not write the DA software, and cannot possibly know why it keeps killing the nameservers."

----------------

The above is what I am up against. Is this a known issue or something that can be fixed? The datacenter was attempting to support DirectAdmin at my request. They normally only offer 'that other control panel' but was going to start offering DA as well if everything went well with mine and on their test server.

Can anyone help? Otherwise its back to CP (and I dont want to do that)

Is it any particular time the problems occur.............. when named gets reloaded for new entries (when a site gets created and dns changes in DA take place)?
 
Hello,

DirectAdmin assums that named is already installed on the server when it's installed. The only commands DA sends to it is:

service named reload

as that's all thats really needed (except manual admin controls).

So, if you run that command, you'll be doing exactly what DA does, so you'll need to keep an eye out for errors to see why that might be affecting things.

John
 
Re: Re: update

ProWebUK said:
Is it any particular time the problems occur.............. when named gets reloaded for new entries (when a site gets created and dns changes in DA take place)?

My admin account is all that is setup so far. Fresh install of RH9 then DA. Each time I went to the box yesterday to begin moving accounts over, nameservers were down.
 
Last edited:
Well, you can see what John has said and theres the confirmation....... its more than likely *not* DA causing the problems unless its the relaods causing a problem and then if service named reload causes problems its more than likely a problem with the application itself.

Chris
 
Back
Top