[Observium] Many false positives

Hi

I'm trying to troubleshoot the many false positives we are receiving from OB.

The system will report a host as down, yet our legacy Nagios and out-of-band Pingdom do not show the host as down.

It doesn't appear that OB records in the log what specifically is making OB think the host is down.

I've increased the SNMP time out value to 3 seconds (which seems very long) and that has helped with some hosts, mostly Mikrotiks.

But I doubt that our Juniper MX480's (which are lightly loaded) should need such long time frame to respond.

How can I get OB to report what is the actual trigger for its "Host Down" alerts ??

Are there tweaks for performance monitoring / testing ??

Thank you in advance..