I’m having a VPS running on OpenVZ that I most of the push to the limit. On usual computers the most significant resource is the memory usage, however on a VPS like this there are a lot of other resources running out.
To get a notion of what happen when the server “crashed”/stopped responding I created a collection of scripts that use the rrdtool to store the statistics and generate nice graphics.
Complete System Monitoring
The program collects all variables found in /proc/user_beancounters and other system parameters.
Requirements
You will need python and rrdtool before running any programs.
Download and Installation
Download using git:
git clone http://src.endnode.se/git/csm
You will get two runnable files, one updating/collecting the data and another for generating the graphics.
Preferably you should make the update script start at system startup and the graph script should be called regularly by a cron daemon but that is not included here.
Perfect work! Thank you very much!
Just one thing:
Would you please advise others that we need a script /root/bin/bean and some sudo-config.
I guess your script has just one line:
———————
#!/bin/sh
cat /proc/user_beancounters
———————
Kind Regards
Christian
Thanks Christian
The package is now updated with the bean.sh file and the additional installation instructions in the README file.
New update available!
Now it’s possible to track memory consumption on a per user basis. Users must however be preconfigured.