URGENT: Please Help, sites down, named screwed up

[nhouse]

Please Help if you possibly can...

I am going to catch the devil from my clients any time now... so if you are able to comment, I need your expertise. Basically here is what gives...

Over the weekend I set up a remote server to backup my domains to. I set up scheduling for the domains, spreading them out over about a two hour time frame so the server wouldn't get blasted by the process. I attempted to backup around 60 domains which included around 5gigs of data. Last night was the first run and this morning I saw that BIND (named) in the Plesk control panel was off... I tried to start it in the panel and got this error message:

ERROR Up LevelUp Level

Unable to make action: Unable to manage service by dnsmng: dnsmng: Service named failed to start

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

0: /usr/local/psa/admin/htdocs/server/restart_services.php:19 psaerror(string "Unable to make action: Unable to manage service by dnsmng: dnsmng: Service named failed to start")


I rebooted the server and the service still showed up as off... however when I logged in with SSH, it appears to be on. I tried restarting it in SSH and I got this error message:

Stopping named: rndc: connect failed: connection refused.
Starting named: [root@nhousemedia httpdocs]# /etc/init.d/ named start
named: already running

Ok, the problem is that many of the domains are not functioning along with their email. Some are though... so I have to assume that somehow named or something associated with it got fubared. I also went back and turned off (disabled) all of the scheduled backups and rebooted again... with no change.

PLEASE help me with some advice if you can... I would sincerely appreciate it! The phone will start ringing soon!

 

[nhouse]

More information from the message logs are below... I hope this will help you to help me... starting to sweat a bit now... already getting phone calls:

Jan 3 01:31:03 nhousemedia named[20672]: loading configuration from '/etc/named.conf'
Jan 3 01:31:03 nhousemedia named[20672]: no IPv6 interfaces found
Jan 3 01:31:03 nhousemedia named[20672]: none:0: open: /etc/rndc.key: file not found
Jan 3 01:31:03 nhousemedia named[20672]: loading configuration from '/etc/named.conf'
Jan 3 01:31:03 nhousemedia named[20672]: no IPv6 interfaces found

Jan 3 05:35:28 nhousemedia named[1905]: zone goodrichtool.com/IN: loading master file goodrichtool.com: file not found
Jan 3 05:35:28 nhousemedia named[1905]: zone hcpbe.com/IN: loading master file hcpbe.com: file not found
Jan 3 05:35:28 nhousemedia named[1905]: zone hickmanrealty.com/IN: loading master file hickmanrealty.com: file not found
Jan 3 05:35:28 nhousemedia named[1905]: zone hillboren.com/IN: loading master file hillboren.com: file not found
Jan 3 05:35:28 nhousemedia named[1905]: zone iis-systems.com/IN: loading master file iis-systems.com: file not found
Jan 3 05:35:28 nhousemedia named[1905]: zone jacksonareabuilders.com/IN: loading master file jacksonareabuilders.com: file not found

HELP!!!

 

[nhouse]

More data from the error_log:

[Tue Jan 03 03:08:09 2006] [warn] RSA server certificate CommonName (CN) `plesk' does NOT match server name!?
[Tue Jan 03 03:08:09 2006] [warn] RSA server certificate is a CA certificate (BasicConstraints: CA == TRUE !?)

HELP!

 

[phoenixisp]

First try this from shell:

#killall -9 named

Then:

#service named start

What happens?

 

[nhouse]

Thanks for the reply... I was in such a bad way that I had to hire someone to assist me. I'll let you know how that goes later.

crossing my fingers...

 

[phoenixisp]

Good luck!!

 

[jshanley]

It appears your mysql database is damaged or not running properly. I'm sure the consultant you hired got it straightened out, but I'm curious to see if I'm close to right. Let us know when you get it resolved

 

[nhouse]

Ok... first of all I want to give a virtual shout out to Cranky for the help... he is the consultant I used to fix my mess and I was under the gun with some of my clients... not fun. It was a paid service and he got to it very quickly... and he has also answered questions here on the forum for me... along with several other wonderful people, I should say. I have been involved in this type of business for a few years now in a small way BUT the is the first time that I have to serve as the server admin. I love the learning BUT hate the stress that you all know when the phone starts ringing off the wall with complaints and you aren't sure what to do. Whew... makes your hair turn gray AND turn loose!

Now, to this point from talking to Cranky I do know that he had to rebuild the DNS zones. I don't know if we can be sure of what caused things to screw up initially. I had used the traditional Plesk backup utility to back up the domains to a remote FTP site on a spread out schedule. Now I am a bit scared of using that again. I can tell you though that I had tested each individual domain to make sure they would connect and backup properly... and they all did. I even manually created a backup of each with no issues.

 

[rvdmeer]

A couple of months ago i also had a strange problem with the named config file. It appeared that plesk did a wrong insert somewere in the middle of the config file. I was able to resolve the problem quickly because i did know where to search.

Anwyho, I dont know if this was the same but maybe cranky could tell us.

Because if it was.. it is a plesk error and it should be submitted.

 

[Cranky]

nhouse,
Thanks for the kind words.

rvdmeer,
No, this was not related. You're referring to a bug in Plesk 7.5.4's earlier builds which the developers now think is fixed (I say 'think' as they aren't 100% sure, but made some changes to avoid the reasons they think most likely caused the problem). Nhouse's dns zones didn't exist at all, they had been deleted from the zones directory, I restored them from Plesk database.

 

[rvdmeer]

Thanks for the reply, its all clear. Cant say anything more then that

 

[jwdick]

Originally posted by rvdmeer
A couple of months ago i also had a strange problem with the named config file. It appeared that plesk did a wrong insert somewere in the middle of the config file. I was able to resolve the problem quickly because i did know where to search.

Anwyho, I dont know if this was the same but maybe cranky could tell us.

Because if it was.. it is a plesk error and it should be submitted.

I have this happen about every other day. Plesk rewrites the named.conf file with errors embedded in the file after my backups run. Now I check the named.conf file every morning to remove the errors in the zone files. What a pain!

 

[Cranky]

What build of Plesk are you using? Have you reported this to SWsoft?

cat /usr/local/psa/version

 

[jwdick]

Originally posted by Cranky
What build of Plesk are you using? Have you reported this to SWsoft?

cat /usr/local/psa/version

7.5.4 RedHat 9 75051115.01

I have not reported this to Sw-Soft. I have not had much luck in my communications with them.

 

[rvdmeer]

They you got an up2date version.
We have; 7.5.4 FedoraCore 3 75051115.01

And i'm not experiencing any problems.
If i was you, i should submit it.

 

[jwdick]

Originally posted by rvdmeer
They you got an up2date version.
We have; 7.5.4 FedoraCore 3 75051115.01

And i'm not experiencing any problems.
If i was you, i should submit it.

Maybe you are right. In the past, SW-Soft has always replied that I need to purchase support from them or use my providers support (GoDaddy). GoDaddy always replies that it's a Plesk bug. Everybody passes the buck!

I'll give it another go with Sw-Soft.

 

[rvdmeer]

I must say that i know of that problem. In the beginning our server provider was Amenworld and when contacting plesk they told us to contact Amenworld because the serial was not known. When contacting Amenworld they told us we should contact plesk. Now with the new provider (FlexServers) they are answering questions regarding plesk bugs..

So.. good luck!

 

[Cranky]

Amen should have been supporting you, not SWsoft.

I've reported this thread to SWsoft to see if the bug can be sorted.

 

[rvdmeer]

Originally posted by Cranky
Amen should have been supporting you, not SWsoft.

I've reported this thread to SWsoft to see if the bug can be sorted.

Hmmm i see.. We where not that happy with Amen because of the support and the connection performance. For example;

The server needed to be reinstalled because of an quota error. (quota's didnt work on a specific kernel release)

Amen support told me that this should be done by clicking on the reinstall server button. They also told me that the reinstall will start 1 minute after i pushed the button.. so i made the backups... When i was done with the backup i pushed the button at 01:00 (night) but nothing happend. 11 hours later the process started...

At that time it was 12:00 and customers where busy with there websites... then suddenly the server restarted and a new install was done.. with the result that Customer lost all there changes they made between 01:00 12:00 because i couldnt make any backups anymore.

The only response was that support didnt know of this...

And there are more stories about Amen performance and support... Ok... sorry for my offtopic story and thanks for sharing

 

[Cranky]

It's basically the responsibility of your reseller to provide support for Plesk, SWsoft will only support direct customers with active support contracts.