grafana: add dashboards and fix permissions
All checks were successful
Run nix flake check / flake-check (push) Successful in 2m3s
All checks were successful
Run nix flake check / flake-check (push) Successful in 2m3s
- Change default OIDC role from Viewer to Editor for Explore access - Add declarative dashboard provisioning - Add node-exporter dashboard (CPU, memory, disk, load, network, I/O) - Add Loki logs dashboard with host/job filters Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
85
services/grafana/dashboards/logs.json
Normal file
85
services/grafana/dashboards/logs.json
Normal file
@@ -0,0 +1,85 @@
|
||||
{
|
||||
"uid": "logs-homelab",
|
||||
"title": "Logs - Homelab",
|
||||
"tags": ["loki", "logs", "homelab"],
|
||||
"timezone": "browser",
|
||||
"schemaVersion": 39,
|
||||
"version": 1,
|
||||
"refresh": "30s",
|
||||
"templating": {
|
||||
"list": [
|
||||
{
|
||||
"name": "host",
|
||||
"type": "query",
|
||||
"datasource": {"type": "loki", "uid": "loki"},
|
||||
"query": "label_values(host)",
|
||||
"refresh": 2,
|
||||
"includeAll": true,
|
||||
"multi": false,
|
||||
"current": {"text": "All", "value": "$__all"}
|
||||
},
|
||||
{
|
||||
"name": "job",
|
||||
"type": "query",
|
||||
"datasource": {"type": "loki", "uid": "loki"},
|
||||
"query": "label_values(job)",
|
||||
"refresh": 2,
|
||||
"includeAll": true,
|
||||
"multi": false,
|
||||
"current": {"text": "All", "value": "$__all"}
|
||||
},
|
||||
{
|
||||
"name": "search",
|
||||
"type": "textbox",
|
||||
"current": {"text": "", "value": ""},
|
||||
"label": "Search"
|
||||
}
|
||||
]
|
||||
},
|
||||
"panels": [
|
||||
{
|
||||
"id": 1,
|
||||
"title": "Log Volume",
|
||||
"type": "timeseries",
|
||||
"gridPos": {"h": 6, "w": 24, "x": 0, "y": 0},
|
||||
"datasource": {"type": "loki", "uid": "loki"},
|
||||
"targets": [
|
||||
{
|
||||
"expr": "sum by (host) (count_over_time({host=~\"$host\", job=~\"$job\"} |~ \"$search\" [1m]))",
|
||||
"legendFormat": "{{host}}",
|
||||
"refId": "A"
|
||||
}
|
||||
],
|
||||
"fieldConfig": {
|
||||
"defaults": {
|
||||
"unit": "short"
|
||||
}
|
||||
},
|
||||
"options": {
|
||||
"legend": {"displayMode": "list", "placement": "bottom"}
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": 2,
|
||||
"title": "Logs",
|
||||
"type": "logs",
|
||||
"gridPos": {"h": 18, "w": 24, "x": 0, "y": 6},
|
||||
"datasource": {"type": "loki", "uid": "loki"},
|
||||
"targets": [
|
||||
{
|
||||
"expr": "{host=~\"$host\", job=~\"$job\"} |~ \"$search\"",
|
||||
"refId": "A"
|
||||
}
|
||||
],
|
||||
"options": {
|
||||
"showTime": true,
|
||||
"showLabels": true,
|
||||
"showCommonLabels": false,
|
||||
"wrapLogMessage": true,
|
||||
"prettifyLogMessage": false,
|
||||
"enableLogDetails": true,
|
||||
"sortOrder": "Descending"
|
||||
}
|
||||
}
|
||||
]
|
||||
}
|
||||
208
services/grafana/dashboards/node-exporter.json
Normal file
208
services/grafana/dashboards/node-exporter.json
Normal file
@@ -0,0 +1,208 @@
|
||||
{
|
||||
"uid": "node-exporter-homelab",
|
||||
"title": "Node Exporter - Homelab",
|
||||
"tags": ["node-exporter", "prometheus", "homelab"],
|
||||
"timezone": "browser",
|
||||
"schemaVersion": 39,
|
||||
"version": 1,
|
||||
"refresh": "30s",
|
||||
"templating": {
|
||||
"list": [
|
||||
{
|
||||
"name": "instance",
|
||||
"type": "query",
|
||||
"datasource": {"type": "prometheus", "uid": "prometheus"},
|
||||
"query": "label_values(node_uname_info, instance)",
|
||||
"refresh": 2,
|
||||
"includeAll": false,
|
||||
"multi": false,
|
||||
"current": {}
|
||||
}
|
||||
]
|
||||
},
|
||||
"panels": [
|
||||
{
|
||||
"id": 1,
|
||||
"title": "CPU Usage",
|
||||
"type": "timeseries",
|
||||
"gridPos": {"h": 8, "w": 12, "x": 0, "y": 0},
|
||||
"datasource": {"type": "prometheus", "uid": "prometheus"},
|
||||
"targets": [
|
||||
{
|
||||
"expr": "100 - (avg by(instance) (rate(node_cpu_seconds_total{mode=\"idle\", instance=~\"$instance\"}[5m])) * 100)",
|
||||
"legendFormat": "CPU %",
|
||||
"refId": "A"
|
||||
}
|
||||
],
|
||||
"fieldConfig": {
|
||||
"defaults": {
|
||||
"unit": "percent",
|
||||
"min": 0,
|
||||
"max": 100,
|
||||
"thresholds": {
|
||||
"mode": "absolute",
|
||||
"steps": [
|
||||
{"color": "green", "value": null},
|
||||
{"color": "yellow", "value": 70},
|
||||
{"color": "red", "value": 90}
|
||||
]
|
||||
}
|
||||
}
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": 2,
|
||||
"title": "Memory Usage",
|
||||
"type": "timeseries",
|
||||
"gridPos": {"h": 8, "w": 12, "x": 12, "y": 0},
|
||||
"datasource": {"type": "prometheus", "uid": "prometheus"},
|
||||
"targets": [
|
||||
{
|
||||
"expr": "(1 - (node_memory_MemAvailable_bytes{instance=~\"$instance\"} / node_memory_MemTotal_bytes{instance=~\"$instance\"})) * 100",
|
||||
"legendFormat": "Memory %",
|
||||
"refId": "A"
|
||||
}
|
||||
],
|
||||
"fieldConfig": {
|
||||
"defaults": {
|
||||
"unit": "percent",
|
||||
"min": 0,
|
||||
"max": 100,
|
||||
"thresholds": {
|
||||
"mode": "absolute",
|
||||
"steps": [
|
||||
{"color": "green", "value": null},
|
||||
{"color": "yellow", "value": 70},
|
||||
{"color": "red", "value": 90}
|
||||
]
|
||||
}
|
||||
}
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": 3,
|
||||
"title": "Disk Usage",
|
||||
"type": "gauge",
|
||||
"gridPos": {"h": 8, "w": 8, "x": 0, "y": 8},
|
||||
"datasource": {"type": "prometheus", "uid": "prometheus"},
|
||||
"targets": [
|
||||
{
|
||||
"expr": "100 - ((node_filesystem_avail_bytes{instance=~\"$instance\",mountpoint=\"/\",fstype!=\"rootfs\"} / node_filesystem_size_bytes{instance=~\"$instance\",mountpoint=\"/\",fstype!=\"rootfs\"}) * 100)",
|
||||
"legendFormat": "Root /",
|
||||
"refId": "A"
|
||||
}
|
||||
],
|
||||
"fieldConfig": {
|
||||
"defaults": {
|
||||
"unit": "percent",
|
||||
"min": 0,
|
||||
"max": 100,
|
||||
"thresholds": {
|
||||
"mode": "absolute",
|
||||
"steps": [
|
||||
{"color": "green", "value": null},
|
||||
{"color": "yellow", "value": 70},
|
||||
{"color": "red", "value": 85}
|
||||
]
|
||||
}
|
||||
}
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": 4,
|
||||
"title": "System Load",
|
||||
"type": "timeseries",
|
||||
"gridPos": {"h": 8, "w": 8, "x": 8, "y": 8},
|
||||
"datasource": {"type": "prometheus", "uid": "prometheus"},
|
||||
"targets": [
|
||||
{
|
||||
"expr": "node_load1{instance=~\"$instance\"}",
|
||||
"legendFormat": "1m",
|
||||
"refId": "A"
|
||||
},
|
||||
{
|
||||
"expr": "node_load5{instance=~\"$instance\"}",
|
||||
"legendFormat": "5m",
|
||||
"refId": "B"
|
||||
},
|
||||
{
|
||||
"expr": "node_load15{instance=~\"$instance\"}",
|
||||
"legendFormat": "15m",
|
||||
"refId": "C"
|
||||
}
|
||||
],
|
||||
"fieldConfig": {
|
||||
"defaults": {
|
||||
"unit": "short"
|
||||
}
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": 5,
|
||||
"title": "Uptime",
|
||||
"type": "stat",
|
||||
"gridPos": {"h": 8, "w": 8, "x": 16, "y": 8},
|
||||
"datasource": {"type": "prometheus", "uid": "prometheus"},
|
||||
"targets": [
|
||||
{
|
||||
"expr": "time() - node_boot_time_seconds{instance=~\"$instance\"}",
|
||||
"legendFormat": "Uptime",
|
||||
"refId": "A"
|
||||
}
|
||||
],
|
||||
"fieldConfig": {
|
||||
"defaults": {
|
||||
"unit": "s"
|
||||
}
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": 6,
|
||||
"title": "Network Traffic",
|
||||
"type": "timeseries",
|
||||
"gridPos": {"h": 8, "w": 12, "x": 0, "y": 16},
|
||||
"datasource": {"type": "prometheus", "uid": "prometheus"},
|
||||
"targets": [
|
||||
{
|
||||
"expr": "rate(node_network_receive_bytes_total{instance=~\"$instance\",device!~\"lo|veth.*|br.*|docker.*\"}[5m])",
|
||||
"legendFormat": "Receive {{device}}",
|
||||
"refId": "A"
|
||||
},
|
||||
{
|
||||
"expr": "-rate(node_network_transmit_bytes_total{instance=~\"$instance\",device!~\"lo|veth.*|br.*|docker.*\"}[5m])",
|
||||
"legendFormat": "Transmit {{device}}",
|
||||
"refId": "B"
|
||||
}
|
||||
],
|
||||
"fieldConfig": {
|
||||
"defaults": {
|
||||
"unit": "Bps"
|
||||
}
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": 7,
|
||||
"title": "Disk I/O",
|
||||
"type": "timeseries",
|
||||
"gridPos": {"h": 8, "w": 12, "x": 12, "y": 16},
|
||||
"datasource": {"type": "prometheus", "uid": "prometheus"},
|
||||
"targets": [
|
||||
{
|
||||
"expr": "rate(node_disk_read_bytes_total{instance=~\"$instance\",device!~\"dm-.*\"}[5m])",
|
||||
"legendFormat": "Read {{device}}",
|
||||
"refId": "A"
|
||||
},
|
||||
{
|
||||
"expr": "-rate(node_disk_written_bytes_total{instance=~\"$instance\",device!~\"dm-.*\"}[5m])",
|
||||
"legendFormat": "Write {{device}}",
|
||||
"refId": "B"
|
||||
}
|
||||
],
|
||||
"fieldConfig": {
|
||||
"defaults": {
|
||||
"unit": "Bps"
|
||||
}
|
||||
}
|
||||
}
|
||||
]
|
||||
}
|
||||
Reference in New Issue
Block a user