Restoring 1 users backup-file, results all other domains showing "Apache is func"etc.

BBM

Verified User
Joined
Jun 8, 2013
Messages
399
Location
Dutch Mountains
Restoring 1 users backup-file, results all other domains showing "Apache is func"etc.

I just did a user_backup restore of 1 user.
Right after that all the other (74) domains are showing "Apache is functioning normally" under their domains.

The only domain showing it's contents correctly now is of that restored user.


In the DA system.log I'm seeing only one thing different than before this restore, and that's all the 'Tally User' commands now have 'mask 20" behind them.

What did I break?
 
I've found it has something to do with Custombuild (2) and mod_ruid(2) perhaps.

Somewhere last week I updated Custombuild to version 2 (wasn't using it before) but still didn't decide to just run it, but instead manually update some services.
With mod_ruid now default ON (I was also not using it as far as I know), and me restoring a users backup today, probably has broken httpd now.
Unfortunatly my limited knowledge doesn't get me any far.

I have just altered the custombuild configuration and turned mod_ruid2 to NO.
But I still get an errormessage after ./build rewrite_confs

Code:
Restarting apache.
Stopping httpd:                                            [FAILED]
Starting httpd: httpd: Syntax error on line 50 of /etc/httpd/conf/httpd.conf: Syntax error on line 1 of /etc/httpd/conf/extra/httpd-phpmodules.conf: Cannot load /usr/lib/apache/mod_mpm_prefork.so into server: /usr/lib/apache/mod_mpm_prefork.so: cannot open shared object file: No such file or directory

Any help availlable to get me up and running again?
 
Was that backup made with custombuild1 configuration or custombuild2 configuratoin?

Maybe something different in saved user-level httpd.conf which then needs to be checked when backups are restored. Caused by mod_ruid2? Apache? I don't know, but I've asked John to look at the thread.

Jeff
 
Hello,

Just as a general followup, if you see the "Apache is functioning normally" page, this is the related debug guide:
http://help.directadmin.com/item.php?id=242

Usually, the command:
Code:
./build rewrite_confs
should re-add the correct configs, if anything is missing.. possibly from /etc/httpd/conf/extra/directadmin-vhosts.conf.

If you've updated to CB2, but are only half using it.. (eg: for some services).. if other services (eg: mod_ruid2) is turned on, but was never actually compiled in, you could run into issues.
Generally:
Code:
./build all d
would be an easy way to ensure everything that is set in the options.conf is actually installed.

John
 
Oops, sorry for the late reply. Didn't expect the topic to come alive again.

The (daily/cron) user_backups at the time were made with CB2 installed.
I tried rebuilding/rewriting the configs a good number of times but everytime httpd just wouldn't start. In the meantime some 75+ domains were down.

Since I don't know my way around CentOS/linux/unix that well, I'm usually very careful not to 'just' try out or do software updates on the VPS.
In the above instance I didn't have a recent VPS-snapshot to fall back on so I wanted to 'fix' the issue by actually reparing what I somehow had wrecked a week ago.
Usually I Google my way through certain issues with some moderate succes. The bigger issues get resolved by RichardG usually :)

I have a feeling my system has a few tiny issues that need to be taken care off, along with a number of (growing) service updates already. But I need to plan a 'nightly' update & test-session for that in case I fubar things, and have a fresh snapshot-backup ready if I get lost all together.
 
But I need to plan a 'nightly' update & test-session for that in case I fubar things, and have a fresh snapshot-backup ready if I get lost all together.
Yep that's a good idea. Just let me know when you have a time table, then we can plan a nightly update together, in case somethings does strange I can instantly have a look at it for you.:)
 
Back
Top