We are currently are trying to use the SNMP Trap receiver Beta to get real time updates from applications when problems arise. Upon testing this we noticed that PRTG doesn't do anything with the received trap until the polling period has been reached.

Why does the Trap receiver have a polling period, a trap should be instantaneous otherwise it's no better than an active polling sensor? Once the trap receiver receives the message it should immediately be acted upon.

Can you explain this behavior or have we misconfiguration something somewhere?

Thanks.

Michael

Article Comments

Dear Michael,

Please consider that the Trap Receiver Sensor is still in BETA state. Every sensor in PRTG has got a scanning interval due to the internal functioning. You can adjust the scanning interval to your needs in the settings of the sensor. Nevertheless I'll forward your request to our wish list for future versions of PRTG.

Best regards


Oct, 2014 - Permalink

Hello,
Speaking about pooling and trap receiving, can you please explain how does it work exactly? Is it possible to "miss" a trap because the pooling interval of the sensor is too big ? Is very any risk to lose traps?

As far as we understand the matter, we have to find a compromise:

  • large pooling interval -> we will know that something happens too late or we can lose the trap
  • small pooling interval -> the alarm is so quick that we do not notice it.

We need the alarm to be up for at least 30 seconds for every trap received and to be sure to receive 100% traps sent.

Thank you for your answer. Best regards


Jul, 2015 - Permalink

Hello ado_byes, your findings are correct.
With the current version of the Syslog and Trap Receivers, you can/must configure a scanning interval, because all sensors within PRTG require a scanning interval.

All messages will be received instantaneously, and be displayed within the sensor (in the messages tab and on the sensor's overview), the scanning interval currently affects the behavior of the Warning Filter and the Error Filter.
Once the scanning interval is over the received messages will be checked for messages matching the warning/error filters, the sensor's state will change and notifications will be processed. The Warning or Error states will be kept until the next scanning interval.

This link describes the sensor's working and behavior:
Blog: Introducing the New High Performance Syslog and SNMP Trap Receiver Sensors

Regarding the possibility of missing traps/syslog messages, that could only happen if the sensor is overloaded with a huge amount of messages at once, otherwise all messages will be received/store, regardless of interval.


Jul, 2015 - Permalink