Backup Failure on 1 of 3 Servers

M

Mitsurugi

Verified User
Joined
Jul 19, 2005
Messages
39
We have 3 Directadmin servers, and all 3 are configured identically. We have the backups scheduled to take place at 3am. We only recently moved from system backups to admin backups for convenience. The backups are failing on just one server.

Here are the facts:

- It's failed for 2 days running, the other 2 servers are working perfectly.

- The backup directory is empty, no files were created. No message was sent via the messaging system (messages are sent on the other 2 working servers).

- A manual backup triggered via the admin console, rather than cron, DOES create the files and sends the message when the job is completed.

- Cron is running, all other cron jobs are run correctly.

- In /var/log/cron.1 today I see this (so it did run):
Aug 5 03:00:00 server1 CROND[17173]: (diradmin) CMD (echo -e "action=backup&id=1&owner=admin" >> /usr/local/directadmin/data/task.queue.da; /bin/chmod 600 /usr/local/directadmin/data/task.queue.da)

- All error.log and errortaskq.log files are empty, no errors reported.

I've run out of ideas. How can I debug this?
 
Hello,

One way is to execute the same echo command manually, then quickly type:
Code: 
/usr/local/directadmin/dataskq d2000
to run the backup in the foreground in debug mode.

Also check /usr/local/directadmin/data/admin/next_ticket.number, to ensure there is a number in it (altough DA should automatically repair itself if thats broken)

In any case, the debug mode should be the best method to see what's going on if the backups are not finishing.

Last thing to check is for any other cronjobs assigned to the task.queue at roughly the same time, and test moving the backup to a later time just to see if that's the cause or not.

John
 
I'm running it now by hand, as you described above, and it's working flawlessly. This means the task scheduler is running, and the job works.

This would suggest that cron didn't trigger the job, but the cron log shows that it did.

The only thing I can think is that I'm logged in as root right now. Could there be a permission issue when the cron job is run? What should the permissions be on the backup directory you specify in the back end?
 
mmmh, usually when a backup fail you receive a ticked on directadmin with the error and what uer backup was doing when error came out, should be disk space, mysql/email using atm, another backup or bad permisss

check it out
 
You must log in or register to reply here.
Back
Top