I just got LibreNMS up and running in my work environment and this is actually the first SNMP monitoring I’ve successfully set up. I’m primarily concerned with monitoring traffic on our 8 switches as we’ve had complaints of “network slowness” for a while now.
I love the graph view as it gives quick visibility into traffic, but after looking at it for the last couple of days, I realize that I don’t know the best way to interpret the data. Logically, I would expect to establish a baseline of data and then any bumps outside that average would be deemed irregular. Aside from that, I’m not sure how I can make better use of the graph data. Any input is welcome.
@dave247 Do you have any ports that are showing errors increasing at the point where the users are complaining?
The usual process for troubleshooting “slowness” is to look carefully at the entire path the user’s data takes step-by-step for errors or things like a port-flap or a mismatched speed/duplex (which can be the result of a failed autonegotiation or possibly a cable fault)
I have one myself where a pair of gig-capable ports at a remote site negotiate a 1G interconnect and then immediately burst into hundreds of thousands of errors per second - it seems the cable between is not truly 1G-capable in reality because it runs infinitely faster when nailed to 100M)
So how does the graph output help? Other than a quick view of whether it’s an uplink overload - the most useful graph is probably the error count graph.
That’s why I said to click on the interface you are troubleshooting it will display all the interface stats including duplex mismatchs along with errors and discards for the interface in question.
