Neil Aggarwal's picture

There may be a memory leak in the Wordpress appliance. 

I downloaded the ISO and installed it to a VPS running on top of a host running CentOS 5.5 KVM.

After turning the site live, everything works fine for a while and then the console shows the kernel is killing processes because the system has run out of memory.  The VPS has 512 MB of RAM allocated to it.  From the webmin usage info, the usage stays at about 300MB for a long time after the appliance is turned on.

This keeps happening so it seems to be a systemic problem.  We tried turning off all plugins to ensure it is not something in the blog and the appliance still locks up.

Since this is a production system, we are moving it to a CentOS based Wordpress install.  If that is stable, I would highly suspect the TKL appliance has some underlying problem.

Has anyone encountered a similar situation?

Any ideas how to diagnose where the problem could be?


Chaim Krause's picture

To test this, you might create a virtual machine and run the appliance as a liveCD. Just do an initial configuration and then let it sit. See if the memory footprint grows. If not, maybe configure another box to continually browse the blog and see if that triggers any growth. This all might be useful data to find the issue.

Liraz Siri's picture

It would be helpful if you could take a closer look and help us diagnose what's happening. For instance, what process is leaking? WordPress? The kernel? Who's using up all the memory? Does restarting the web server help? etc.
Neil Aggarwal's picture

The CentOS VPS is doing the same thing so am I now thinking TKL is not the problem.  Sorry for the false alarm.

Add new comment