I have a single Linux server and it's occasionally unreachable for a while. I want to know what causes this (I suspect either high CPU or Memory usage). I would like to monitor:
- Global CPU/Memory usage
- CPU/Memory usage of individual processes (/etc/init.d services would suffice)
- Network traffic (broken down by process/service would be neat)
I'd like to see a graphs, but I'm not afraid to generate them myself using GNUPlot.
The server resources are limited, so I don't want to use any heavyweight tools.
On Windows machines, there is typically a tool called perfmon that can do this - is there any resource-saving tool like that for Linux? All I know is Nagios - too heavy weight for my case?