27 lines
974 B
Nix
27 lines
974 B
Nix
{ lib, pkgs, config, ... }: # Standard module arguments
|
|
{
|
|
# This module contributes its rule group to a list that will be
|
|
# collected and processed by the main vmalert module.
|
|
services.vmalert.rules.groups = [
|
|
{
|
|
name = "CPUUsageAlerts";
|
|
# interval = "60s"; # Optional: group-level interval
|
|
rules = [ # This MUST be a list of rule attribute sets
|
|
{
|
|
alert = "HighCPUUsage";
|
|
expr = "(1 - avg by (instance, job) (rate(node_cpu_seconds_total{mode=\"idle\"}[1m]))) * 100 > 90";
|
|
for = "5m";
|
|
labels = {
|
|
severity = "warning";
|
|
category = "performance";
|
|
};
|
|
annotations = {
|
|
summary = "High CPU usage on {{ $labels.instance }}";
|
|
description = "CPU usage on {{ $labels.instance }} (job: {{ $labels.job }}) has been above 90% for more than 5 minutes. Current value: {{ $value | printf \"%.2f\" }}%.";
|
|
};
|
|
}
|
|
];
|
|
}
|
|
];
|
|
}
|