Management IP and SNMP Unavailable behavior

[horse1bun]

We have several Netonix across our network. We are seeing in our monitoring that the devices become unavailable to ping and SNMP is showing the devices are not responding for intermittent intervals. This behavior is consistent across our various devices. The switches do not stop passing traffic, just the management response seems to be affected.

I searched the forums but don't see any similar issues reported. I am wondering if anyone else is seeing this and if there is something we are doing wrong?

We use Mikrotik routers with vlans to our various netowrk segments and a management vlan for the switch. In this is example, ports 25 and 26 are in LAG but this is not always the case.

It doesn't seem as though we are maxing the resources of any of the switches.

[sirhc]

Why do you have port 25,26 checked as Trunks, are you passing more VLANs than you have defined in your VLAN Matrix?

[horse1bun]

On this particular switch, we have a LAG setup for 25+26 into our CCR 1036 on a bonding interface and have assigned our Vlans to this bonding interface to pass Vlan traffic through both ports for highest possible future capacity.

We only intend to use the Vlans listed on the page. However, we have been allowing all Vlans 1-4095 on T ports as we are still learning best practices.
Reading the Netonix forums, it would seem you recommend best practice on the T ports is that we should limit the allowed Vlans, in this example, to "901-923, 4000"?

[horse1bun]

Unless I need to allow all / other Vlans through the Trunk ports, I might try having it setup more like this?

[dshef]

Could we figure out why these devices are having such large gaps in the graphs? We have several graphs in libreNMS, all have several gaps throughout the day, constantly...snmp polling is timing out or something? Is this a firmware bug or normal behavior?

Does anybody else see large gaps in graphing in LibreNMS?