Hello,
this is explained in the sensor itself (PRTG web interface, with enabled help):
What is this?
The System, Core, and Probe Health sensors monitor internal PRTG parameters and are created automatically by PRTG. Various parameters of your PRTG system are checked which can affect the quality of the monitoring results:
- Core CPU Load: Extensive CPU load can lead to false, incomplete and even incorrect monitoring results. Should usually stay below 50%.
- Core Handles: Counter for data structures of the operating system, responsible for internal resource management. Repeated obviously increasing values should be investigated.
- Core Memory Usage: Amount of memory used by the PRTG core server as reported by the memory manager. Repeated obviously increasing values should be investigated. Constantly high values above 2 GB indicate that PRTG runs at its limits.
- Core Threads: Number of program parts currently running simultaneously. This value can increase with heavy load.
- Health: This is an index value which sums up the probe state into a value between 100% (healthy) and 0% (failing). Frequent values below 100% should be investigated.
- Probe CPU Load: Extensive CPU load can lead to false, incomplete and even incorrect monitoring results. Should usually stay below 50%.
- Probe Handles: Counter for data structures of the operating system, responsible for internal resource management. Repeated obviously increasing values should be investigated.
- Probe Interval Delay non-WMI&SNMP: This is the interval delay value for all other sensor types.
- Probe Interval Delay SNMP: This is the interval delay value for SNMP sensors. If this is above 0% there are probably very slow SNMP V3 sensors. Try to increase the monitoring intervals or distribute the sensors over several probes.
- Probe Interval Delay WMI: This value is above 0% if WMI sensors could not be checked according to their interval. 100% means that WMI sensors on the average are checked with twice their interval. For values above 0% try to increase the monitoring intervals or distribute the sensors over several probes to keep the number of WMI sensors per probe below 120 (with 60s interval) or 600 (with 300s interval).
- Probe Memory Usage: Amount of memory used by the PRTG probe service as reported by the memory manager. Repeated obviously increasing values should be investigated. Constantly high values above 2 GB indicate that PRTG runs at its limits (and you should distribute some sensors to Remote Probes).
- Probe Message Queue: This is the number of monitoring results from the probe not yet processed by the core. Should usually stay below 1/10 of the sensor count.
- Probe Open Requests: Number of currently active monitoring requests. Should stay below the maximum of 500.
- Probe Threads: Number of program parts currently running simultaneously. This value can increase with heavy monitoring load.
- System Available Memory: Amount of free memory available on the system. This value should not fall below 500 MB, so PRTG can still request recources during report generation, auto-discoveries, etc.
- System Available Memory Percent: Free memory available on the system in percent.
- System CPU Load: Extensive CPU load on the system (e.g. caused by other running applications) can lead to false, incomplete and even incorrect monitoring results. Should usually stay below 50%.
See Also
Add comment