LXC Monitoring collects cpu and memory statistics for several Linux Containers, and creates pretty charts with it. Although it is intended for LXC containers, it can be used to monitor any type of cgroup. Optionally it can be configured to send a daily/weekly/... mail with those charts.
I'm a Development Environments guy and we've been quite impressed with your LXC-Monitor package, which I am currently using on a test environment to monitor a number of LXC containers. My question is, however, that it would be a huge advantage for me to be able to generate the charts using percentages for RAM and CPU usage. Obviously, to achieve this the system needs to know the maximum amount of RAM/CPU for a container, but I believe this could be done by the data_poller.py taking the relevant CGROUP data from the same directory as it looks for each container's current usage. I would be extremely grateful if you can be of any help, and I'd obviously be willing to give you recognition on our system.
Traceback (most recent call last):
File "generate_charts.py", line 190, in
main()
File "generate_charts.py", line 179, in main
generateCPUChart(c, group, starttime, endtime, args.folder)
File "generate_charts.py", line 91, in generateCPUChart
g('set xtics %d ' % ( (diff[-1][0]-diff[0][0])/8, ))
IndexError: list index out of range