Go to All Forums

How to monitor the nproc limit?

Hi Team,

I have a server running Oracle WCS and have set "nproc" limits in /etc/security/limits.d/20-nproc.conf.  So, my question is how do I monitor the nproc limits?

How do I make sure that the process should not fall under a certain limit If that happens how do we set a threshold and send email alerts?

In other words, how to handle the monitoring and be alerted when you close the fixed limit?

Thanks in advance!!

 

Thanks,

Swapnil Pawar

 

Like (2) Reply
Replies (3)

Hi Swapnil Pawar,

Thank you for asking. 

With Site24x7 Custom Plugins, users can write their own plugin and monitor the "nproc" limits. 

Please refer to the below link to write a custom plugin: 

https://www.site24x7.com/help/admin/adding-a-monitor/plugins/custom-plugins.html

After creating the custom plugin, the user will get options to set the threshold for the "nproc" metric and get alerted if the threshold is breached. 

For further queries, please reply to this thread. 

Thanks,

Muralikrishnan,
Product Manager,
Site24x7 

Like (0) Reply

Hi Swapnil Pawar,

Greetings! 

We are happy to inform you that we have implemented an exclusive plugin to monitor user process count. You can set threshold limits and get alerts if the user process count exceeds a particular threshold.

What is nproc?

nproc is a built-in command in a Linux system that is used to count the number of processing units available in a current system or to a current process. The maximum user processes (nproc) limit on Linux is the number of threads within all processes that can exist for a particular user. 1024 is the default value of nproc on some versions of Linux, and it is generally an insufficient number of threads for all processes. Technicians get error messages like "resource temporarily unavailable" or "error creating thread" when nproc is too low.

To avoid these errors because of low nproc, technicians can use the plugin available in the below link and monitor user processes count:

https://github.com/site24x7/plugins/tree/master/process_count_monitoring

Plugin installation 

Create a directory named "process_count_monitoring" under the Site24x7 Linux Agent plugin directory:

  Linux    ->   /opt/site24x7/monagent/plugins/process_count_monitoring

Download all the files in the "process_count_monitoring" folder and place them under the "process_count_monitoring" directory.

Execute the below command with appropriate arguments to check for the valid json output:

python3 process_count_monitoring.py --user_name=<USER NAME> 

Configuration

Provide the username in the process_count_monitoring.cfg file to monitor the user processes count. 

In the next data collection event of the, the plugin will be discovered and marked up for monitoring. You can see the monitor under Server > Plugin Integrations. The plugin monitor will also be listed under the respective server monitor's Plugins tab (Server > Server Monitor > Servers > click on the desired server monitor > Plugins). You can also set up threshold profiles and get alerts when the configured value exceeds set limits.

If you have any queries, please reply to this thread.

Regards,

Muralikrishnan,
Product Manager,
Site24x7

Like (0) Reply

Hey Murali, Thank you for your contribution. We have already built our own nproc plugin (similar to what you have developed) and It's running and monitoring our resource properly.

 

Thanks,

Swapnil Pawar

Like (0) Reply

Was this post helpful?