We use Solarwinds to monitor IP SLA performance on some important MPLS connections.
We have them set up to alert us if a circuit starts responding at slower than 160 ms.
An email went get sent to Tier 1 support to investigate.
You can read how to set up alerts on Solarwinds: https://support.solarwinds.com/Success_Center/Server_Application_Monitor_(SAM)/Create_a_component_alert
Now, A slow ping every once in a while is expected and fine. We don’t want to cry wolf. To keep the noise level down and actually only raise an incident when the line is degraded we need to implement alert timers.
There are 3 timers we can use in a Solarwinds Alert:
1. Trigger Condition Timer
How long an issue needs to be reported before an alert is raised.
Adjust this to make the alert more or less sensitive.
2. Reset Condition Timer
How long an issue needs to stop happening before the alert is automatically cleared.
Adjust this to keep the alert from flapping. you don’t want a million emails on a link that keeps going up/down.
3. Triger Action Timer
How long the alert needs to be active before the action happens (usually an email, but can do a lot of things.)
Adjust this to give the link a chance to stabalize. Set it longer than the reset timer (or 2x the reset timer depending on urgency)
There are also polling timers that come in to play, but we don’t generally alter them.