Portal Home > Knowledgebase > Industry Announcements > Web Hosting Main Forums > Providers and Network Outages and Updates > How to monitor processes that are killed?
How to monitor processes that are killed?
| Posted by Lakjin, 10-26-2016, 02:49 AM |
|
Recently I moved my dedicated server to CloudLinux. After this move, some processes appear to be randomly stopping. I'm wondering if there is a CloudLinux or Apache setting that is killing these processes. How can I monitor which processes are killed by CloudLinux and/or Apache?
Thanks!
|
| Posted by victor_y, 10-26-2016, 03:06 AM |
|
You can get some info in the error log, but it depends on the apache debug level exactly what types of errors are logged there.
Check out this page for more info on the subject: https://httpd.apache.org/docs/2.4/logs.html
|
| Posted by nixaserver, 10-26-2016, 05:17 AM |
|
Selinux enabled by default in server and some times restrict processes.The default location where you can find this logging is /var/log/audit/audit.log.
|
| Posted by Lakjin, 10-26-2016, 02:37 PM |
|
Thank you both, will have a look.
|
| Posted by helpdesk, 10-26-2016, 03:27 PM |
|
Usually if OS kills processes due to resource overwhelmed, it logs that termination message in kernel logs /var/log/messages or debug command. You can also read apache logs with warn / debug level enabled.
|
| Posted by zoid, 10-26-2016, 03:45 PM |
|
Which processes and why would you think it is Apache? Neither of them should stop any processes, particularly not Apache. You would need to be more specific in order to get anything remote to a useful answer.
|
| Posted by Srv24x7, 10-27-2016, 01:24 AM |
|
Hi,
You will have to monitor the /var/log/messages log to check for events like this. Check if it is processes are killed because of memory issues or something else. This will be the first point in the direction of investigation.
|
| Posted by Bazingaaa, 10-28-2016, 10:18 AM |
|
Can you please specify what exactly processes were killed? CloudLinux could kill 'hunged' php processes if their parent pid is 1. Also, if users are reaching their memory limit - kernel will kill process and write something to /var/log/messages. Useful command to check faults for a last day:
lveinfo --by-fault any --period=1d
<>
Last edited by anon-e-mouse; 11-10-2016 at 05:59 AM.
|
Add to Favourites 
Print this Article
Also Read
Shared Hosting
Shared Hosting
