To keep track of the hardware and network resources consumed during their operation, Nexthink Appliances internally run collectd.
The collectd daemon gathers the following performance metrics by default:
aggregation: Sum and average of the usage over all CPUs.
cpu: Amount of time spent by the CPU in various states; namely executing user code, executing system code, waiting for IO-operations and being idle.
df: File system usage information; basically how much space on a mounted partition is used and how much is available. Named after the df(1) UNIX command and very similar to it.
disk: Performance statistics of hard-disks and, where supported, partitions.
interface: Information about the traffic (bytes, packets and errors per second) of interfaces.
load: System load. Rough overview over the usage of a machine.
memory: Physical memory usage as reported by the operating system.
ntpd: Drifts of the ntpd daemon from the reference NTP server.
processes: Process-related information about the Engine and the Portal.
swap: Swap usage.
vmem: Virtual memory usage.
Getting performance data
To get the RRD files generated by collectd:
Log in to the Web Console of the Appliance to monitor (Portal or Engine).
Select the APPLIANCE tab at the top of the Web Console.
Click Monitoring on the left-hand side menu.
Find the Performance data section:
Click DOWNLOAD. Your web browser downloads a compressed tape archive file (tar.gz extension) that holds the collectd performance data and includes the current date and time in its name.