[Panzer Max]

Well, it shouldn't get stuck, but I think that's what happened.

cpuwatch (Thu Dec 31 01:00:03 2015): System load is currently 9.17; waiting for it to go down to 7.00 to continue …
====

From the above errors, it is clear that the backup process was terminated because of high load in the server. After this, I have checked the load statistics of the server. I have found that the server load is high and it is nearly 16 and there are high accesses to the server.

===
root@server [~]# w
15:18:46 up 898 days, 22:16, 1 user, load average: 16.52, 16.25, 16.12
USER TTY FROM LOGIN@ IDLE JCPU PCPU WHAT
root pts/0 61.15.45.80 14:36 0.00s 0.08s 0.00s w

root@server [~]# nproc
8
root@server [~]#

root@server [~]# netstat -plane | grep :80 | wc -l
620
====

I have also analyzed the previous load statistics and could see that the load was high for the past days. The average server load is nearly 16, which is double the threshold value. We infer that the server load was always 16, after Dec 31, 2015 and this is the reason for backup failure.

And inlinemod.php was the culprit. There must be something to explain this....

Can you give me some examples of exactly what "inline moderation" is?

Edit: http://www.vbulletin.com/forum/forum...certain-timing

I guess it's simply hitting the delete or edit key on a thread/post? Maybe selecting a group of threads to move....