Hello,

I'm monitoring a HP VirtualConnect and all of it's bandwidth interfaces. However, the virtual connect is terrible at replying every time PRTG scans the sensor. I've set the scanning interval to "Set sensor to warning for 3 interval, then set to down" in order to avoid false "error" alarms.

However, this behavior still generates tons of log entries, as in "New log entries" to the left of the overall sensor status in the top right corner. In matter of hours there are thousands of new log entries, due to the sensors of the virtual connect switching back and forth from OK to Warning.

Is there any way that I can disable new log entries for a particular device? Perhaps there should be another option under "Scanning interval" which says something like "Set sensor to OK for 3 interval, then to warning".

Kind regards, Thomas

Article Comments

Hello Thomas,
thank you for your KB-Post.

At the moment there's no plan to implement such change. How many SNMP Sensors do you have deployed against this device and in which scanning interval?

You might be able to "fine tune" PRTG's queries by modifying the Credentials for SNMP Devices, SNMP Compatibility Options and Scanning Interval.

You might also want to check with HP if there is a newer firmware/software version available for the VirtualConnect, that may improve the performance of the SNMP Agent. On some devices there are specific commands or options to raise the priority of SNMP, which usually eliminates this sort of errors.

Best Regards,
Luciano Lingnau [Paessler Support]


Sep, 2016 - Permalink

Hello Luciano,

I really appreciate your reply.

We currently have 48 SNMP traffic sensors in total to the two VirtualConnects in High Availability. I will try and increase the SNMP Delay to allow the VC to have more time to reply. However, it even doesn't reply to all Ping packets we send, so I'm not sure increasing the SNMP Delay will fix anything. I will remove the limit on "Packet Loss" on our Ping sensors as well.

It's not critical to gather this information every 60 seconds, so we don't wish to down prioritize SNMP compared to other critical features. That's why it would be sweet to have PRTG allow a couple of missed replies without going into warning state and just use PRTG for statistic purposes for this particular device.

Let's see if increasing SNMP delay and remove packet loss limit on the ping sensors will help!

Best regards,
Thomas


Sep, 2016 - Permalink

Unfortunately it hasn't made a difference. They still go into warning state and produce a log entry.

If possible, I'd like to make a feature request of turning off log entry creations for certain devices.

Thank you!

Best regards, Thomas


Sep, 2016 - Permalink

Hello Thomas,
thank you for your reply.

Please review the following KB-Post to understand how we rate customer feature requests:

I don't want to disappoint you but a lot of users would have to request this feature for it to have ana real chance of being implemented anytime soon. As of now our perception is that if something like this happens this is an issue that should be addressed on the monitored device and not in PRTG.

Best Regards,
Luciano Lingnau [Paessler Support]


Sep, 2016 - Permalink

Hello Luciano,

I totally understand. Was worth a shot! :-)

Best regards, Thomas


Sep, 2016 - Permalink