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>
1027 lines
22 KiB
JSON
1027 lines
22 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": "Service 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=~\"ml-training.*\", 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=~\"ml-training.*\"})",
|
|
"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=~\"ml-training.*\", 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": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(container_memory_working_set_bytes{namespace=\"foxhunt\", pod=~\"ml-training.*\", container!=\"\"}) / sum(kube_pod_container_resource_limits{namespace=\"foxhunt\", pod=~\"ml-training.*\", resource=\"memory\"})",
|
|
"legendFormat": "",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Memory",
|
|
"type": "gauge"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${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": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(container_memory_working_set_bytes{namespace=\"foxhunt\", pod=~\"ml-training.*\", container!=\"\"})",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Memory (bytes)",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${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": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(rate(container_cpu_usage_seconds_total{namespace=\"foxhunt\", pod=~\"ml-training.*\", container!=\"\"}[5m]))",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "CPU Cores",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${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": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(rate(container_network_receive_bytes_total{namespace=\"foxhunt\", pod=~\"ml-training.*\"}[5m]))",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Network RX",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${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": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "sum(rate(container_network_transmit_bytes_total{namespace=\"foxhunt\", pod=~\"ml-training.*\"}[5m]))",
|
|
"instant": true,
|
|
"legendFormat": "",
|
|
"range": false,
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Network TX",
|
|
"type": "stat"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 5
|
|
},
|
|
"id": 10,
|
|
"panels": [],
|
|
"title": "Resources",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${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": 6
|
|
},
|
|
"id": 11,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(container_cpu_usage_seconds_total{namespace=\"foxhunt\", pod=~\"ml-training.*\", container!=\"\"}[$interval])",
|
|
"legendFormat": "{{container}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "CPU Usage",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
},
|
|
"unit": "bytes"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 8,
|
|
"y": 6
|
|
},
|
|
"id": 12,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "container_memory_working_set_bytes{namespace=\"foxhunt\", pod=~\"ml-training.*\", container!=\"\"}",
|
|
"legendFormat": "{{container}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Memory",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
},
|
|
"unit": "Bps"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 8,
|
|
"x": 16,
|
|
"y": 6
|
|
},
|
|
"id": 13,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(container_network_receive_bytes_total{namespace=\"foxhunt\", pod=~\"ml-training.*\"}[$interval])",
|
|
"legendFormat": "RX {{pod}}",
|
|
"refId": "A"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(container_network_transmit_bytes_total{namespace=\"foxhunt\", pod=~\"ml-training.*\"}[$interval])",
|
|
"legendFormat": "TX {{pod}}",
|
|
"refId": "B"
|
|
}
|
|
],
|
|
"title": "Network I/O",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 14
|
|
},
|
|
"id": 14,
|
|
"panels": [],
|
|
"title": "Job Management",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 12,
|
|
"x": 0,
|
|
"y": 15
|
|
},
|
|
"id": 15,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "kube_pod_status_phase{namespace=\"foxhunt\", pod=~\"training-.*\", phase=\"Running\"}",
|
|
"legendFormat": "{{pod}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Active Training Jobs",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 12,
|
|
"x": 12,
|
|
"y": 15
|
|
},
|
|
"id": 16,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "kube_pod_status_phase{namespace=\"foxhunt\", pod=~\"training-.*\", phase=\"Succeeded\"}",
|
|
"legendFormat": "{{pod}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Job Completions",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 23
|
|
},
|
|
"id": 17,
|
|
"panels": [],
|
|
"title": "gRPC",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
}
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 12,
|
|
"x": 0,
|
|
"y": 24
|
|
},
|
|
"id": 18,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "rate(grpc_server_handled_total{pod=~\"ml-training.*\"}[$interval])",
|
|
"legendFormat": "{{grpc_method}}",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "gRPC Handled",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"fieldConfig": {
|
|
"defaults": {
|
|
"custom": {
|
|
"drawStyle": "line",
|
|
"fillOpacity": 10,
|
|
"lineWidth": 1,
|
|
"stacking": {
|
|
"mode": "none"
|
|
}
|
|
},
|
|
"unit": "s"
|
|
},
|
|
"overrides": []
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 12,
|
|
"x": 12,
|
|
"y": 24
|
|
},
|
|
"id": 19,
|
|
"options": {
|
|
"legend": {
|
|
"displayMode": "list",
|
|
"placement": "bottom"
|
|
},
|
|
"tooltip": {
|
|
"mode": "multi",
|
|
"sort": "desc"
|
|
}
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "prometheus",
|
|
"uid": "${DS_PROMETHEUS}"
|
|
},
|
|
"expr": "histogram_quantile(0.99, rate(grpc_server_handling_seconds_bucket{pod=~\"ml-training.*\"}[$interval]))",
|
|
"legendFormat": "p99",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "gRPC Latency",
|
|
"type": "timeseries"
|
|
},
|
|
{
|
|
"collapsed": false,
|
|
"gridPos": {
|
|
"h": 1,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 32
|
|
},
|
|
"id": 20,
|
|
"panels": [],
|
|
"title": "Logs",
|
|
"type": "row"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "loki",
|
|
"uid": "${DS_LOKI}"
|
|
},
|
|
"gridPos": {
|
|
"h": 10,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 33
|
|
},
|
|
"id": 21,
|
|
"options": {
|
|
"dedupStrategy": "none",
|
|
"enableLogDetails": true,
|
|
"prettifyLogMessage": false,
|
|
"showCommonLabels": false,
|
|
"showLabels": true,
|
|
"showTime": true,
|
|
"sortOrder": "Descending",
|
|
"wrapLogMessage": true
|
|
},
|
|
"targets": [
|
|
{
|
|
"datasource": {
|
|
"type": "loki",
|
|
"uid": "${DS_LOKI}"
|
|
},
|
|
"expr": "{namespace=\"foxhunt\", pod=~\"ml-training.*\"} | pattern `<_> <_> <_> <content>` | regexp `msg=\"(?P<clean_msg>[^\"]*)\"` | line_format `{{if .clean_msg}}{{.clean_msg}}{{else}}{{.content}}{{end}}` | drop content, clean_msg",
|
|
"refId": "A"
|
|
}
|
|
],
|
|
"title": "Log Stream",
|
|
"type": "logs"
|
|
},
|
|
{
|
|
"datasource": {
|
|
"type": "loki",
|
|
"uid": "${DS_LOKI}"
|
|
},
|
|
"gridPos": {
|
|
"h": 8,
|
|
"w": 24,
|
|
"x": 0,
|
|
"y": 43
|
|
},
|
|
"id": 22,
|
|
"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=~\"ml-training.*\"} | json | 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": [
|
|
"service",
|
|
"ml",
|
|
"training"
|
|
],
|
|
"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": "ML Training Service",
|
|
"uid": "svc-ml-training"
|
|
}
|