torjus/nixos-servers
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects 1 Releases Wiki Activity

grafana: fix datasource UIDs for VictoriaMetrics migration
Some checks failed
Run nix flake check / flake-check (push) Has been cancelled
Details

Browse Source 
Update all dashboard datasource references from "prometheus" to
"victoriametrics" to match the declared datasource UID. Enable
prune and deleteDatasources to clean up the old Prometheus
(monitoring01) datasource from Grafana's database.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Torjus Håkestad
2026-02-17 22:23:04 +01:00
parent 95a96b2192
commit 65acf13e6f
8 changed files with 90 additions and 86 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
services/grafana/dashboards/systemd.json
View File

@@ -15,7 +15,7 @@
{
"name": "hostname",
"type": "query",
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"query": "label_values(systemd_unit_state, hostname)",
"refresh": 2,
"includeAll": true,
@@ -30,7 +30,7 @@
"title": "Failed Units",
"type": "stat",
"gridPos": {"h": 4, "w": 4, "x": 0, "y": 0},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "count(systemd_unit_state{state=\"failed\", hostname=~\"$hostname\"} == 1) or vector(0)",
@@ -60,7 +60,7 @@
"title": "Active Units",
"type": "stat",
"gridPos": {"h": 4, "w": 4, "x": 4, "y": 0},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "count(systemd_unit_state{state=\"active\", hostname=~\"$hostname\"} == 1)",
@@ -86,7 +86,7 @@
"title": "Hosts Monitored",
"type": "stat",
"gridPos": {"h": 4, "w": 4, "x": 8, "y": 0},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "count(count by (hostname) (systemd_unit_state{hostname=~\"$hostname\"}))",
@@ -112,7 +112,7 @@
"title": "Total Service Restarts",
"type": "stat",
"gridPos": {"h": 4, "w": 4, "x": 12, "y": 0},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "sum(systemd_service_restart_total{hostname=~\"$hostname\"})",
@@ -143,7 +143,7 @@
"title": "Inactive Units",
"type": "stat",
"gridPos": {"h": 4, "w": 4, "x": 16, "y": 0},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "count(systemd_unit_state{state=\"inactive\", hostname=~\"$hostname\"} == 1)",
@@ -169,7 +169,7 @@
"title": "Timers",
"type": "stat",
"gridPos": {"h": 4, "w": 4, "x": 20, "y": 0},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "count(systemd_timer_last_trigger_seconds{hostname=~\"$hostname\"})",
@@ -195,7 +195,7 @@
"title": "Failed Units",
"type": "table",
"gridPos": {"h": 6, "w": 12, "x": 0, "y": 4},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "systemd_unit_state{state=\"failed\", hostname=~\"$hostname\"} == 1",
@@ -251,7 +251,7 @@
"title": "Service Restarts (Top 15)",
"type": "table",
"gridPos": {"h": 6, "w": 12, "x": 12, "y": 4},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "topk(15, systemd_service_restart_total{hostname=~\"$hostname\"} > 0)",
@@ -309,7 +309,7 @@
"title": "Active Units per Host",
"type": "bargauge",
"gridPos": {"h": 8, "w": 12, "x": 0, "y": 10},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "sort_desc(count by (hostname) (systemd_unit_state{state=\"active\", hostname=~\"$hostname\"} == 1))",
@@ -339,7 +339,7 @@
"title": "NixOS Upgrade Timers",
"type": "table",
"gridPos": {"h": 8, "w": 12, "x": 12, "y": 10},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "systemd_timer_last_trigger_seconds{name=\"nixos-upgrade.timer\", hostname=~\"$hostname\"}",
@@ -429,7 +429,7 @@
"title": "Backup Timers",
"type": "table",
"gridPos": {"h": 6, "w": 12, "x": 0, "y": 18},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "systemd_timer_last_trigger_seconds{name=~\"restic.*\", hostname=~\"$hostname\"}",
@@ -524,7 +524,7 @@
"title": "Service Restarts Over Time",
"type": "timeseries",
"gridPos": {"h": 6, "w": 12, "x": 12, "y": 18},
"datasource": {"type": "prometheus", "uid": "prometheus"},
"datasource": {"type": "prometheus", "uid": "victoriametrics"},
"targets": [
{
"expr": "sum by (hostname) (increase(systemd_service_restart_total{hostname=~\"$hostname\"}[1h]))",