We have some devices for which the manufacturer has advised too many snmp queries will cause a crash.
To help prevent this, is there any way for PRTG to limit the number of snmp queries per second sent to a particular device? For example, we might have hundreds or thousands of sensors on this device, but we only want 20 queries per second max to go to the device, so PRTG should spread out the queries over time.
Article Comments
Thank you, I think that delay will do exactly what I was hoping.
Can you describe what would happen if the scanning interval comes up again before the first scanning interval is completed? So if a scan takes 6 minutes to complete and the interval is for 5 minutes, does a second scan kick off at the 5 minute mark? Or does the second scan get delayed by a minute?
Dec, 2018 - Permalink
That won't happen, since it will run into a timeout first (as configured in the device' SNMP Compatibility Settings tab), which is defaulting to five seconds. Should you decide to set the interval to something higher, make sure that the timeout is smaller than the interval (5m interval, 4m timeout). Otherwise, the requests might cause the device to crash :)
PRTG Scheduler |
PRTGapi |
Feature Requests |
WMI Issues |
SNMP Issues
Kind regards,
Stephan Linke, Tech Support Team
Dec, 2018 - Permalink
There's no such option, unfortunately. You could however increase the SNMP Delay option (up to 100ms) in the device settings, under "SNMP Compatibility Options". Another way to decrease the amount of requests would be to use a large interval (i.e. every five minutes, or even 10, if the devices are ever so fragile).
PRTG Scheduler | PRTGapi | Feature Requests | WMI Issues | SNMP Issues
Kind regards,
Stephan Linke, Tech Support Team
Dec, 2018 - Permalink