Skip to content

Metrics#

Metrics Collected#

In addition to executing plans, the client also does some basic metric collection. The main categories that are collected are:

System Info#

The client collects the following system information:

  • OS
  • OS type
  • Kernel
  • Hostname**
  • IP address
  • Uptime
  • Date client joined
  • Location
  • Host type

**The hostname is discovered from the host, not from DNS, so in environments like AWS your DNS FQDN and your local hostname my not match.

CPU#

  • Model
  • Speed
  • Cache size
  • Cores
  • Usage data (Used %)

Memory#

  • Usage data (Total, Free, Used, %)
  • Swap usage data (Total, Free, Used, %)

Disk#

  • Paritions
  • Device name/path
  • Filesystem
  • Mount point
  • Usage data (Total, Free, used, %)
  • Inode usage data (Total, Free, Used, %)

Network#

  • Number interfaces
  • Interface name
  • MAC address
  • IPv4 address
  • IPv6 address
  • Interface performance metrics (In, Out, Err)
  • Network Connections
    • Listen address
    • Remote port
    • Local port
    • PID

Process#

  • Number of processes
  • PID
  • Name
  • Status
  • Path
  • Arguments
  • Username
  • Parent process
  • Child processes
  • Thread count
  • CPU & Memory usage (Total, Free, Used, %)

Docker#

By default the client will discover and interface with a local Docker installation and report metrics and info about the Docker server and its running containers. This data consists of the following metric points for each container

  • Name
  • IP address
  • Container status
  • Creation time
  • Start time
  • Host that container is running on
  • Image
  • Ports
  • Command
  • Restart policy
  • Environment variables**
  • Lables
  • Mounts
  • Networks

**Reminder that environment variables are not secure and should not be used to pass secrets to a container

Packages#

  • Package manager
  • Package name
  • Package version

A note about data collection#

It is occasionally necessary to collect more data points than are presented to the user. The data presented here represents the totality of metrics collected, not necessarily metrics presented.

We will never collect or send metrics elsewhere without your explicit agreement; this includes any kind of usage or performance metrics. By default we do not collect usage and performance metrics from on-premise installations. In fact, our software does not "phone home" for any reason as we believe that your data should stay that way. For debugging purposes we may ask you for log files or application output, but this will be explicit and entirely up to the user to provide.