I have a customer who has started using virtualmin pro and his server keeps going berserk and locking up due to excessively high load. We are slowly working out the kinks but I saw something strange and I am unable to determine why it is happening.
When I start a service like HTTPD or NAMED using the standard startup scripts in CentOS 5 it takes an incredibly long time. When I look at the process list to see what is going on using 'ps afx' it shows me the following
/bin/bash - /etc/init.d/named start - /bin/unicode_start xxxxxxxxxxxxxx - /bin/unicode_start xxxxxxxxxxxxxxx
The unicode_start line continues on and on and on and on with another one nested below the other. There were a whole bunch of these so many that it goes way off the console and shows what appear to be blank lines. I piped the output to less and could scroll over and indeed there were a whole bunch.
Does anybody have any idea why it would start so many of these and if this is why things are taking so long to load?
That's weird. I have no idea what it is, but none of my systems exhibit this behavior...I'd want to suggest that maybe it's some sort of library linking problem, but it's so far outside of anything I've ever seen, I'd just be guessing.
What OS/version is this? Is it a virtualized system, perhaps using a CoW filesystem? (I'm grasping for something to hang a theory on...)
--
Check out the forum guidelines!