PRTG Manual: Object Hierarchy
All objects in the PRTG monitoring configuration are arranged in a tree-like hierarchy. You can arrange the objects in groups that monitor similar devices, services, or particular locations. This hierarchical order is also used to define common settings for larger groups of objects. The settings of the root group, for example, apply to all other objects underneath in the hierarchy by default (see section Inheritance of Settings).
The graph shows the object hierarchy in PRTG:
- The root group contains all objects in your setup. All probes are located underneath the root group.
- A probe contains one or several groups.
- A group contains one or several devices.
- A device represents a physical or virtual component in your network that is reachable via an IP address. Several sensors belong to one device.
- A sensor monitors one single aspect of a device and has at least one channel.
- A channel tracks one specific data stream and displays it as part of a sensor.
The root group is the topmost instance in PRTG. It contains all other objects in your setup. We recommend that you adjust all default settings for the root group. This is because all other objects in the device tree inherit these standard settings by default. Thus, you do not have to set up the same configuration for each object anew.
Each group (except the root group) is part of a probe. This is the platform where the actual monitoring takes place. All objects that you add to a probe are monitored via that probe. For PRTG on premises, every PRTG core server installation automatically installs the local probe. For PRTG hosted by Paessler, every instance comes with the hosted probe.
You can add additional probes and remote probes to your configuration to also monitor remote devices from outside your network (see section Multiple Probes and Remote Probes). In a cluster, there is an additional cluster probe that runs on all cluster nodes. Devices on the cluster probe are monitored by all cluster nodes, so that monitoring always continues, even if one of the cluster nodes fails.
On each probe, there are one or more groups that have structural purposes. Use groups to arrange similar objects so that they inherit the same settings. To a group, you add devices. You can arrange your devices in different nested groups to reflect the structure of your network.
Here is a sample configuration of a device tree with the local probe, several groups, devices, and their sensors.
You can add devices that you want to monitor to each probe or group. Each device in your configuration represents real hardware or a virtual device in your network. This can be, for example:
- Web or file servers
- Client computers (Windows, Linux, or macOS)
- Routers or network switches
- Almost every device in your network that has its own IP address
Sometimes you may want to add the same device several times to receive a better overview when using a large number of sensors for very detailed monitoring, or to use different device settings for different groups of sensors. You can add multiple devices with the same IP address or Domain Name System (DNS) name. The sensors on all of these devices then query the same real hardware device in your network.
PRTG additionally adds the probe device to the local probe. This is an internal system device with several sensors. It has access to the probe system and monitors the system's health parameters.
PRTG automatically analyzes the devices that you add and recommends appropriate sensors on the device's Overview tab. In the Recommended Sensors table, click the Add These Sensors button in the corresponding table row to create recommended sensors with one click.
You can turn off the sensor recommendation in System Administration—Monitoring.
On each device, you can create a number of sensors. Every sensor monitors one single aspect of a device. This can be, for example:
- A network service like Simple Mail Transfer Protocol (SMTP), File Transfer Protocol (FTP), HTTP, and others
- The traffic of a network switch
- The CPU load of a device
- The memory load of a device
- The traffic on one network card
- A NetFlow device
- The system health of a device
- Other content (for example, of databases, mails, HTTP, Extensible Markup Language (XML), files, etc.)
For more information, see the video tutorial: What Is a Sensor?
Every sensor has a number of channels through which it receives the different data streams. The available channels depend on the type of sensor. One channel can contain, for example:
- Downtime and uptime of a device
- Traffic in of a bandwidth device (for example, a router)
- Traffic out of a bandwidth device (for example, a router)
- Traffic sum of a bandwidth device (for example, a router)
- WWW traffic of a NetFlow device
- Mail traffic of a NetFlow device
- Other traffic of a NetFlow device
- CPU load of a device
- Loading time of a web page
- Download bandwidth of a web page
- Time to first byte of a web page
- Response time of a Ping request to a device
- Response time of a Remote Desktop service
What Is a Sensor?