Why is my webserver restarting multiple times a day without my permission?

My client recently switched servers and landed on a server with nodeworx/siteworx installed on it.
Ever since sever migration it has been a daily fight for survival to try and get things to operate as expected on the server.

It is very common that we will ssh into the server to make changes to apache .ini files or mysql .cnf files, get the changes to take effect, then find later that the files have been reverted to their previous configurations. Does Nodeworx/Siteworx have some sort of service that is caching the content of the files and restoring them to their previous configurations on a schedule?

Additionally, and even more importantly, every single day since the migration of the server, services that should never, ever, have any downtime whatsoever on a webserver are being asked to restart and causing outages across our ecosystem. After spending a lot of time digging through logs and running through all other options, we have come to guess that the iworx.pex cron tasks are likely the culprit.

Can anyone advise as to whether you know what the issue may be, and advise what changes need to be made to stop any unexpected and unauthorized changes from happening to our server that are not directly initiated by us? It is frightening going to sleep with something akin to a tamper monkey making changes to the server, updating things without permission, changing config files, restarting apache and mysql and different security services.

Hi

Welcome to IW forums

You do not show what IW version or distro

Firstly InterWorx would not make any changes to your files with exception if IW was updated. Then IW files would be updated so any changes made to IW files may change unless you added the changes to their own files and added in

Is Apache set to soft restart

If your not too sure I’d open a support ticket and let IW have a look to see what’s going on but I think it’s Apache soft restart

Many thanks

John

Hi,

Do you have csf firewall in your server?
If yes, include “exe:/usr/sbin/rotatelogs” in /etc/csf/csf.pignore and restart lfd.

Joao


You wouldn’t happen to be talking about this option that says “Auto-restart Apache” would you?

Yes we do have csf firewall. We have done as you said, hope that it resolves our problem.

Or this “Force Graceful Restart” option? It really sucks that these are mostly unlabelled. Surely “Force Graceful Restart” doesn’t mean, yes please restart the HTTPD service multiple times a day for seemingly no reason, right?

Hi

Yes that what it is called

Apache does need to be restarted when any changes take place such as new SSL etc

Many thanks

John

Hello–

The Autorestart feature just tries to restart Apache automatically if it ever goes down. So if, for instance, there is an OOM issue on the server that takes down Apache–once that is cleared, Apache will turn itself back on, vs you needing to manually restart it after resolving the memory issue.

The Force Graceful Restart setting controls how InterWorx restarts Apache when that occurs (so, for instance, when a new SiteWorx account is created Apache is automatically restarted so that the newlly created vhost goes into effect). That being set to “on” means that InterWorx will call a graceful restart, vs a hard kill.

Details can be found here: Web Server Guide — InterWorx documentation

Neither of those options restart Apache multiple times a day. And neither of them do the hard kills that you provided in the logging snippirt in the ticket you submitted. I responded to your ticket 20 or so minutes ago–once I have access to the server, I should be able to tell you more. :slight_smile:

Thanks,
-Jenna
Friendly Neighborhood InterWorx Support Manager

Wonderful, that is what i was hoping that those features did.