feat: add smart alerting and noatime to disks
This commit is contained in:
102
hosts/web-arm/modules/grafana/alerting/storage/raid_alerts.nix
Normal file
102
hosts/web-arm/modules/grafana/alerting/storage/raid_alerts.nix
Normal file
@@ -0,0 +1,102 @@
|
||||
{ lib, pkgs, config, ... }:
|
||||
{
|
||||
grafanaAlertRuleDefinitions = [
|
||||
# RAID array degraded - critical
|
||||
{
|
||||
uid = "raid-array-degraded-uid";
|
||||
title = "RaidArrayDegraded";
|
||||
condition = "D";
|
||||
data = [
|
||||
{
|
||||
refId = "A";
|
||||
datasourceUid = "vm-datasource-uid";
|
||||
relativeTimeRange = { from = 300; to = 0; };
|
||||
model = {
|
||||
expr = ''mdadm_array_state == 0'';
|
||||
instant = false;
|
||||
};
|
||||
}
|
||||
{
|
||||
refId = "C";
|
||||
datasourceUid = "__expr__";
|
||||
model = {
|
||||
type = "reduce";
|
||||
expression = "A";
|
||||
reducer = "last";
|
||||
};
|
||||
}
|
||||
{
|
||||
refId = "D";
|
||||
datasourceUid = "__expr__";
|
||||
model = {
|
||||
type = "math";
|
||||
expression = "$C == 0";
|
||||
};
|
||||
}
|
||||
];
|
||||
for = "0s";
|
||||
noDataState = "NoData";
|
||||
execErrState = "Error";
|
||||
annotations = {
|
||||
summary = "RAID array {{ $labels.array }} is degraded";
|
||||
description = ''
|
||||
RAID array {{ $labels.array }} on {{ $labels.instance }} is in state "{{ $labels.state }}".
|
||||
The array is not in a healthy state. Check for failed disks immediately!
|
||||
'';
|
||||
};
|
||||
labels = {
|
||||
severity = "critical";
|
||||
category = "storage";
|
||||
};
|
||||
}
|
||||
|
||||
# RAID missing devices - critical
|
||||
{
|
||||
uid = "raid-missing-devices-uid";
|
||||
title = "RaidMissingDevices";
|
||||
condition = "D";
|
||||
data = [
|
||||
{
|
||||
refId = "A";
|
||||
datasourceUid = "vm-datasource-uid";
|
||||
relativeTimeRange = { from = 300; to = 0; };
|
||||
model = {
|
||||
expr = ''mdadm_array_devices_active < mdadm_array_devices_total'';
|
||||
instant = false;
|
||||
};
|
||||
}
|
||||
{
|
||||
refId = "C";
|
||||
datasourceUid = "__expr__";
|
||||
model = {
|
||||
type = "reduce";
|
||||
expression = "A";
|
||||
reducer = "last";
|
||||
};
|
||||
}
|
||||
{
|
||||
refId = "D";
|
||||
datasourceUid = "__expr__";
|
||||
model = {
|
||||
type = "math";
|
||||
expression = "$C > 0";
|
||||
};
|
||||
}
|
||||
];
|
||||
for = "0s";
|
||||
noDataState = "NoData";
|
||||
execErrState = "Error";
|
||||
annotations = {
|
||||
summary = "RAID array {{ $labels.array }} has missing devices";
|
||||
description = ''
|
||||
RAID array {{ $labels.array }} on {{ $labels.instance }} has fewer active devices than expected.
|
||||
A disk may have failed or been removed. Check array status immediately!
|
||||
'';
|
||||
};
|
||||
labels = {
|
||||
severity = "critical";
|
||||
category = "storage";
|
||||
};
|
||||
}
|
||||
];
|
||||
}
|
||||
Reference in New Issue
Block a user