725fefe565
- Add 4 Prometheus alerts: HeadscaleDown (critical), NoOnlineNodes, HighHTTPLatency, HighErrorRate - Add Grafana dashboard with node count, map responses, HTTP latency, nodestore operations, and memory panels - Pin headscale-ui to digest sha256:015f5ba0... (was :latest) - Set disable_check_updates: true to skip GitHub check on startup - Uptime Kuma monitor already existed (id=19, 300s interval)
78 lines
3.1 KiB
JSON
78 lines
3.1 KiB
JSON
{
|
|
"annotations": { "list": [] },
|
|
"editable": true,
|
|
"fiscalYearStartMonth": 0,
|
|
"graphTooltip": 1,
|
|
"links": [],
|
|
"panels": [
|
|
{
|
|
"title": "Online Nodes",
|
|
"type": "stat",
|
|
"gridPos": { "h": 4, "w": 6, "x": 0, "y": 0 },
|
|
"targets": [{ "expr": "headscale_nodestore_nodes_total", "legendFormat": "Nodes" }],
|
|
"fieldConfig": { "defaults": { "thresholds": { "steps": [{ "color": "red", "value": 0 }, { "color": "green", "value": 1 }] } } }
|
|
},
|
|
{
|
|
"title": "Map Responses / sec",
|
|
"type": "timeseries",
|
|
"gridPos": { "h": 8, "w": 9, "x": 6, "y": 0 },
|
|
"targets": [
|
|
{ "expr": "rate(headscale_mapresponse_sent_total[5m])", "legendFormat": "sent" },
|
|
{ "expr": "rate(headscale_mapresponse_generated_total[5m])", "legendFormat": "generated" },
|
|
{ "expr": "rate(headscale_mapresponse_ended_total[5m])", "legendFormat": "ended" }
|
|
]
|
|
},
|
|
{
|
|
"title": "Endpoint Updates / sec",
|
|
"type": "stat",
|
|
"gridPos": { "h": 4, "w": 6, "x": 0, "y": 4 },
|
|
"targets": [{ "expr": "rate(headscale_mapresponse_endpoint_updates_total[5m])", "legendFormat": "updates/s" }],
|
|
"fieldConfig": { "defaults": { "unit": "ops" } }
|
|
},
|
|
{
|
|
"title": "HTTP Request Rate by Path",
|
|
"type": "timeseries",
|
|
"gridPos": { "h": 8, "w": 9, "x": 15, "y": 0 },
|
|
"targets": [{ "expr": "sum by (path) (rate(headscale_http_requests_total[5m]))", "legendFormat": "{{ path }}" }]
|
|
},
|
|
{
|
|
"title": "HTTP p95 Latency by Path",
|
|
"type": "timeseries",
|
|
"gridPos": { "h": 8, "w": 12, "x": 0, "y": 8 },
|
|
"targets": [{ "expr": "histogram_quantile(0.95, sum by (path, le) (rate(headscale_http_duration_seconds_bucket[5m])))", "legendFormat": "{{ path }}" }],
|
|
"fieldConfig": { "defaults": { "unit": "s" } }
|
|
},
|
|
{
|
|
"title": "NodeStore Operations / sec",
|
|
"type": "timeseries",
|
|
"gridPos": { "h": 8, "w": 12, "x": 12, "y": 8 },
|
|
"targets": [
|
|
{ "expr": "rate(headscale_nodestore_operations_total[5m])", "legendFormat": "operations" },
|
|
{ "expr": "headscale_nodestore_queue_depth", "legendFormat": "queue depth" }
|
|
]
|
|
},
|
|
{
|
|
"title": "NodeStore Batch Duration p95",
|
|
"type": "timeseries",
|
|
"gridPos": { "h": 8, "w": 12, "x": 0, "y": 16 },
|
|
"targets": [{ "expr": "histogram_quantile(0.95, rate(headscale_nodestore_batch_duration_seconds_bucket[5m]))", "legendFormat": "p95" }],
|
|
"fieldConfig": { "defaults": { "unit": "s" } }
|
|
},
|
|
{
|
|
"title": "Memory Usage",
|
|
"type": "timeseries",
|
|
"gridPos": { "h": 8, "w": 12, "x": 12, "y": 16 },
|
|
"targets": [
|
|
{ "expr": "go_memstats_alloc_bytes{job=\"kubernetes-service-endpoints\", namespace=\"headscale\"}", "legendFormat": "alloc" },
|
|
{ "expr": "go_memstats_sys_bytes{job=\"kubernetes-service-endpoints\", namespace=\"headscale\"}", "legendFormat": "sys" }
|
|
],
|
|
"fieldConfig": { "defaults": { "unit": "bytes" } }
|
|
}
|
|
],
|
|
"schemaVersion": 39,
|
|
"tags": ["headscale", "vpn"],
|
|
"templating": { "list": [] },
|
|
"time": { "from": "now-6h", "to": "now" },
|
|
"title": "Headscale VPN",
|
|
"uid": "headscale-vpn"
|
|
}
|