dba50fa3be
Update all dashboard JSON files to use Helm chart's Prometheus datasource UID and add Grafana sidecar folder annotations for auto-provisioning. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
1633 lines
35 KiB
JSON
1633 lines
35 KiB
JSON
{
|
|
"annotations": {
|
|
"list": [
|
|
{
|
|
"builtIn": 1,
|
|
"datasource": {
|
|
"type": "grafana",
|
|
"uid": "-- Grafana --"
|
|
},
|
|
"enable": true,
|
|
"hide": true,
|
|
"iconColor": "rgba(0, 211, 255, 1)",
|
|
"name": "Annotations & Alerts",
|
|
"type": "dashboard"
|
|
}
|
|
]
|
|
},
|
|
"editable": true,
|
|
"fiscalYearStartMonth": 0,
|
|
"graphTooltip": 1,
|
|
"links": [],
|
|
"panels": [
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 0
|
|
},
|
|
"id": 1,
|
|
"panels": [],
|
|
"title": "Argo Workflows Health",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "red",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "green",
|
|
"value": 1
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 4,
|
|
"w": 3,
|
|
"x": 0,
|
|
"y": 1
|
|
},
|
|
"id": 2,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "kube_pod_status_phase{namespace=\"foxhunt\", pod=~\"argo-workflows.*\", phase=\"Running\"}",
|
|
"instant": true,
|
|
"legendFormat": "Running",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Pod Status",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "yellow",
|
|
"value": 3
|
|
},
|
|
{
|
|
"color": "red",
|
|
"value": 10
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 4,
|
|
"w": 3,
|
|
"x": 3,
|
|
"y": 1
|
|
},
|
|
"id": 3,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(kube_pod_container_status_restarts_total{namespace=\"foxhunt\", pod=~\"argo-workflows.*\"})",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Restarts",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"max": 1,
|
|
"min": 0,
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "yellow",
|
|
"value": 0.7
|
|
},
|
|
{
|
|
"color": "red",
|
|
"value": 0.9
|
|
}
|
|
]
|
|
},
|
|
"unit": "percentunit"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 3,
|
|
"x": 6,
|
|
"y": 1
|
|
},
|
|
"id": 4,
|
|
"options": {
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(rate(container_cpu_usage_seconds_total{namespace=\"foxhunt\", pod=~\"argo-workflows.*\", container!=\"\"}[$interval]))",
|
|
"legendFormat": "",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "CPU Usage",
|
|
"type": "gauge"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"max": 1,
|
|
"min": 0,
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "yellow",
|
|
"value": 0.7
|
|
},
|
|
{
|
|
"color": "red",
|
|
"value": 0.9
|
|
}
|
|
]
|
|
},
|
|
"unit": "percentunit"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 3,
|
|
"x": 9,
|
|
"y": 1
|
|
},
|
|
"id": 5,
|
|
"options": {
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(container_memory_working_set_bytes{namespace=\"foxhunt\", pod=~\"argo-workflows.*\", container!=\"\"}) / sum(kube_pod_container_resource_limits{namespace=\"foxhunt\", pod=~\"argo-workflows.*\", resource=\"memory\"})",
|
|
"legendFormat": "",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Memory",
|
|
"type": "gauge"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"unit": "bytes"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 4,
|
|
"w": 3,
|
|
"x": 12,
|
|
"y": 1
|
|
},
|
|
"id": 6,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(container_memory_working_set_bytes{namespace=\"foxhunt\", pod=~\"argo-workflows.*\", container!=\"\"})",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Memory (bytes)",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"unit": "short"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 4,
|
|
"w": 3,
|
|
"x": 15,
|
|
"y": 1
|
|
},
|
|
"id": 7,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(rate(container_cpu_usage_seconds_total{namespace=\"foxhunt\", pod=~\"argo-workflows.*\", container!=\"\"}[5m]))",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "CPU Cores",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"unit": "Bps"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 4,
|
|
"w": 3,
|
|
"x": 18,
|
|
"y": 1
|
|
},
|
|
"id": 8,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(rate(container_network_receive_bytes_total{namespace=\"foxhunt\", pod=~\"argo-workflows.*\"}[5m]))",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Network RX",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
},
|
|
"unit": "Bps"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 4,
|
|
"w": 3,
|
|
"x": 21,
|
|
"y": 1
|
|
},
|
|
"id": 9,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(rate(container_network_transmit_bytes_total{namespace=\"foxhunt\", pod=~\"argo-workflows.*\"}[5m]))",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Network TX",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 5
|
|
},
|
|
"id": 10,
|
|
"panels": [],
|
|
"title": "Workflow Metrics",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 4,
|
|
"x": 0,
|
|
"y": 6
|
|
},
|
|
"id": 11,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "argo_workflows_gauge{status=\"Running\"}",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Active Workflows",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 4,
|
|
"x": 4,
|
|
"y": 6
|
|
},
|
|
"id": 12,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "PBFA97CFB590B2093"
|
|
},
|
|
"expr": "argo_workflows_gauge{status=\"Succeeded\"}",
|
|
"legendFormat": "",
|
|
"refId": "A",
|
|
"instant": true,
|
|
"range": false
|
|
}
|
|
],
|
|
"title": "Succeeded",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
},
|
|
{
|
|
"color": "red",
|
|
"value": 1
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 4,
|
|
"x": 8,
|
|
"y": 6
|
|
},
|
|
"id": 13,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "PBFA97CFB590B2093"
|
|
},
|
|
"expr": "argo_workflows_gauge{status=\"Failed\"}",
|
|
"legendFormat": "",
|
|
"refId": "A",
|
|
"instant": true,
|
|
"range": false
|
|
}
|
|
],
|
|
"title": "Failed",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 4,
|
|
"x": 12,
|
|
"y": 6
|
|
},
|
|
"id": 14,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "PBFA97CFB590B2093"
|
|
},
|
|
"expr": "argo_workflows_pods_total_count{phase=\"Pending\"}",
|
|
"legendFormat": "",
|
|
"refId": "A",
|
|
"instant": true,
|
|
"range": false
|
|
}
|
|
],
|
|
"title": "Pending Pods",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 4,
|
|
"x": 16,
|
|
"y": 6
|
|
},
|
|
"id": 15,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "argo_workflows_workers_busy_count",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Workers Busy",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"color": {
|
|
"mode": "thresholds"
|
|
},
|
|
"thresholds": {
|
|
"mode": "absolute",
|
|
"steps": [
|
|
{
|
|
"color": "green",
|
|
"value": null
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 5,
|
|
"w": 4,
|
|
"x": 20,
|
|
"y": 6
|
|
},
|
|
"id": 16,
|
|
"options": {
|
|
"colorMode": "value",
|
|
"graphMode": "area",
|
|
"reduceOptions": {
|
|
"calcs": [
|
|
"lastNotNull"
|
|
]
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "argo_workflows_queue_depth_gauge",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Queue Depth",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 11
|
|
},
|
|
"id": 17,
|
|
"panels": [],
|
|
"title": "Performance",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 0,
|
|
"y": 12
|
|
},
|
|
"id": 18,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "argo_workflows_gauge{status=\"Running\"}",
|
|
"legendFormat": "Running",
|
|
"refId": "A"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "argo_workflows_gauge{status=\"Pending\"}",
|
|
"legendFormat": "Pending",
|
|
"refId": "B"
|
|
}
|
|
],
|
|
"title": "Workflow Count",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 8,
|
|
"y": 12
|
|
},
|
|
"id": 19,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(argo_workflows_k8s_request_total[$interval])",
|
|
"legendFormat": "{{verb}} {{kind}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "K8s API Rate",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
},
|
|
"unit": "s"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 16,
|
|
"y": 12
|
|
},
|
|
"id": 20,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "histogram_quantile(0.99, rate(argo_workflows_operation_duration_seconds_bucket[$interval]))",
|
|
"legendFormat": "p99",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Operation Duration",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
},
|
|
"unit": "s"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 0,
|
|
"y": 20
|
|
},
|
|
"id": 21,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "histogram_quantile(0.99, rate(argo_workflows_queue_latency_bucket[$interval]))",
|
|
"legendFormat": "p99",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Queue Latency",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 8,
|
|
"y": 20
|
|
},
|
|
"id": 22,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(argo_workflows_error_count[$interval])",
|
|
"legendFormat": "errors/s",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Errors",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 16,
|
|
"y": 20
|
|
},
|
|
"id": 23,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "argo_workflows_pods_total_count",
|
|
"legendFormat": "{{status}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Pods Total",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 28
|
|
},
|
|
"id": 24,
|
|
"panels": [],
|
|
"title": "Go Runtime",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 0,
|
|
"y": 29
|
|
},
|
|
"id": 25,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "argo_workflows_process.runtime.go.goroutines",
|
|
"legendFormat": "goroutines",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Goroutines",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
},
|
|
"unit": "bytes"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 8,
|
|
"y": 29
|
|
},
|
|
"id": 26,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "argo_workflows_process.runtime.go.mem.heap_inuse",
|
|
"legendFormat": "in-use",
|
|
"refId": "A"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "argo_workflows_process.runtime.go.mem.heap_alloc",
|
|
"legendFormat": "alloc",
|
|
"refId": "B"
|
|
}
|
|
],
|
|
"title": "Heap Usage",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 16,
|
|
"y": 29
|
|
},
|
|
"id": 27,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(argo_workflows_process.runtime.go.gc.pause_total_ns[$interval])",
|
|
"legendFormat": "ns/s",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "GC Pauses",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 37
|
|
},
|
|
"id": 28,
|
|
"panels": [],
|
|
"title": "Resources",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
},
|
|
"unit": "short"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 0,
|
|
"y": 38
|
|
},
|
|
"id": 29,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(container_cpu_usage_seconds_total{namespace=\"foxhunt\", pod=~\"argo-workflows.*\", container!=\"\"}[$interval])",
|
|
"legendFormat": "{{container}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "CPU Usage",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
},
|
|
"unit": "bytes"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 8,
|
|
"y": 38
|
|
},
|
|
"id": 30,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "container_memory_working_set_bytes{namespace=\"foxhunt\", pod=~\"argo-workflows.*\", container!=\"\"}",
|
|
"legendFormat": "{{container}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Memory",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
},
|
|
"unit": "Bps"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 16,
|
|
"y": 38
|
|
},
|
|
"id": 31,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(container_network_receive_bytes_total{namespace=\"foxhunt\", pod=~\"argo-workflows.*\"}[$interval])",
|
|
"legendFormat": "RX {{pod}}",
|
|
"refId": "A"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(container_network_transmit_bytes_total{namespace=\"foxhunt\", pod=~\"argo-workflows.*\"}[$interval])",
|
|
"legendFormat": "TX {{pod}}",
|
|
"refId": "B"
|
|
}
|
|
],
|
|
"title": "Network I/O",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 46
|
|
},
|
|
"id": 32,
|
|
"panels": [],
|
|
"title": "Logs",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "loki",
|
|
"uid": "${DS_LOKI}"
|
|
},
|
|
"gridPos": {
|
|
"h": 10,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 47
|
|
},
|
|
"id": 33,
|
|
"options": {
|
|
"dedupStrategy": "none",
|
|
"enableLogDetails": true,
|
|
"prettifyLogMessage": false,
|
|
"showCommonLabels": false,
|
|
"showLabels": true,
|
|
"showTime": true,
|
|
"sortOrder": "Descending",
|
|
"wrapLogMessage": true
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "loki",
|
|
"uid": "${DS_LOKI}"
|
|
},
|
|
"expr": "{namespace=\"foxhunt\", pod=~\"argo-workflows.*\"} | pattern `<_> <_> <_> <content>` | regexp `msg=\"(?P<clean_msg>[^\"]*)\"` | line_format `{{if .clean_msg}}{{.clean_msg}}{{else}}{{.content}}{{end}}` | drop content, clean_msg",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Log Stream",
|
|
"type": "logs"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "loki",
|
|
"uid": "${DS_LOKI}"
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 57
|
|
},
|
|
"id": 34,
|
|
"options": {
|
|
"dedupStrategy": "none",
|
|
"enableLogDetails": true,
|
|
"prettifyLogMessage": false,
|
|
"showCommonLabels": false,
|
|
"showLabels": true,
|
|
"showTime": true,
|
|
"sortOrder": "Descending",
|
|
"wrapLogMessage": true
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "loki",
|
|
"uid": "${DS_LOKI}"
|
|
},
|
|
"expr": "{namespace=\"foxhunt\", pod=~\"argo-workflows.*\"} | logfmt | level =~ \"error|warn|fatal\" | pattern `<_> <_> <_> <content>` | regexp `msg=\"(?P<clean_msg>[^\"]*)\"` | line_format `{{if .clean_msg}}{{.clean_msg}}{{else}}{{.content}}{{end}}` | drop content, clean_msg",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Errors & Warnings",
|
|
"type": "logs"
|
|
}
|
|
],
|
|
"refresh": "30s",
|
|
"schemaVersion": 39,
|
|
"tags": [
|
|
"platform",
|
|
"argo",
|
|
"ci"
|
|
],
|
|
"templating": {
|
|
"list": [
|
|
{
|
|
"auto": false,
|
|
"current": {
|
|
"text": "5m",
|
|
"value": "5m"
|
|
},
|
|
"name": "interval",
|
|
"options": [
|
|
{
|
|
"selected": false,
|
|
"text": "1m",
|
|
"value": "1m"
|
|
},
|
|
{
|
|
"selected": true,
|
|
"text": "5m",
|
|
"value": "5m"
|
|
},
|
|
{
|
|
"selected": false,
|
|
"text": "15m",
|
|
"value": "15m"
|
|
},
|
|
{
|
|
"selected": false,
|
|
"text": "30m",
|
|
"value": "30m"
|
|
},
|
|
{
|
|
"selected": false,
|
|
"text": "1h",
|
|
"value": "1h"
|
|
}
|
|
],
|
|
"query": "1m,5m,15m,30m,1h",
|
|
"type": "interval"
|
|
}
|
|
]
|
|
},
|
|
"time": {
|
|
"from": "now-1h",
|
|
"to": "now"
|
|
},
|
|
"timezone": "browser",
|
|
"title": "Argo Workflows",
|
|
"uid": "plat-argo"
|
|
}
|