The purpose is to be able to graph resource usage on on-field device.
resources are polled every 60s to keep logs size low.
log files are compressed/rotated and push to central server
polling information are directly exported in Graphite format.
appart from pure resource polling (such as CPU/RAM/Disk/FileSize/Swap), has been added some nginx usage stats (active connections/total number of connection/etc.)
The purpose is to be able to graph resource usage on on-field device.