{ "__inputs": [], "__requires": [ { "type": "grafana", "id": "grafana", "name": "Grafana", "version": "11.0.0" }, { "type": "datasource", "id": "prometheus", "name": "Prometheus", "version": "1.0.0" }, { "type": "panel", "id": "timeseries", "name": "Time series", "version": "" }, { "type": "panel", "id": "stat", "name": "Stat", "version": "" }, { "type": "panel", "id": "table", "name": "Table", "version": "" } ], "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": 0, "id": null, "links": [], "panels": [ { "datasource": { "type": "prometheus", "uid": "prometheus" }, "fieldConfig": { "defaults": { "color": { "mode": "thresholds" }, "thresholds": { "mode": "absolute", "steps": [ { "color": "green", "value": null }, { "color": "orange", "value": 50 }, { "color": "red", "value": 150 } ] }, "unit": "reqps" }, "overrides": [] }, "gridPos": { "h": 5, "w": 6, "x": 0, "y": 0 }, "id": 1, "options": { "colorMode": "background", "graphMode": "area", "justifyMode": "auto", "reduceOptions": { "calcs": [ "lastNotNull" ], "fields": "", "values": false }, "textMode": "auto" }, "targets": [ { "editorMode": "code", "expr": "sum(job:ml_inference_rps:rate5m{model_name=~\"$model_name\",model_version=~\"$model_version\"})", "legendFormat": "RPS", "range": true, "refId": "A" } ], "title": "Inference Requests / Sec", "type": "stat" }, { "datasource": { "type": "prometheus", "uid": "prometheus" }, "fieldConfig": { "defaults": { "color": { "mode": "thresholds" }, "thresholds": { "mode": "absolute", "steps": [ { "color": "green", "value": null }, { "color": "orange", "value": 0.25 }, { "color": "red", "value": 0.35 } ] }, "unit": "s" }, "overrides": [] }, "gridPos": { "h": 5, "w": 6, "x": 6, "y": 0 }, "id": 2, "options": { "colorMode": "background", "graphMode": "area", "justifyMode": "auto", "reduceOptions": { "calcs": [ "lastNotNull" ], "fields": "", "values": false }, "textMode": "auto" }, "targets": [ { "editorMode": "code", "expr": "max(job:ml_inference_latency_p95_seconds:5m{model_name=~\"$model_name\",model_version=~\"$model_version\"})", "legendFormat": "p95", "range": true, "refId": "A" } ], "title": "P95 Inference Latency", "type": "stat" }, { "datasource": { "type": "prometheus", "uid": "prometheus" }, "fieldConfig": { "defaults": { "color": { "mode": "thresholds" }, "thresholds": { "mode": "absolute", "steps": [ { "color": "green", "value": null }, { "color": "orange", "value": 0.01 }, { "color": "red", "value": 0.03 } ] }, "unit": "percentunit" }, "overrides": [] }, "gridPos": { "h": 5, "w": 6, "x": 12, "y": 0 }, "id": 3, "options": { "colorMode": "background", "graphMode": "area", "justifyMode": "auto", "reduceOptions": { "calcs": [ "lastNotNull" ], "fields": "", "values": false }, "textMode": "auto" }, "targets": [ { "editorMode": "code", "expr": "max(job:ml_inference_error_rate:ratio5m{model_name=~\"$model_name\",model_version=~\"$model_version\"})", "legendFormat": "error_rate", "range": true, "refId": "A" } ], "title": "Inference Error Rate", "type": "stat" }, { "datasource": { "type": "prometheus", "uid": "prometheus" }, "fieldConfig": { "defaults": { "color": { "mode": "thresholds" }, "thresholds": { "mode": "absolute", "steps": [ { "color": "green", "value": null }, { "color": "orange", "value": 1 }, { "color": "red", "value": 3 } ] } }, "overrides": [] }, "gridPos": { "h": 5, "w": 6, "x": 18, "y": 0 }, "id": 4, "options": { "colorMode": "background", "graphMode": "area", "justifyMode": "auto", "reduceOptions": { "calcs": [ "lastNotNull" ], "fields": "", "values": false }, "textMode": "auto" }, "targets": [ { "editorMode": "code", "expr": "max(job:ml_drift_features_triggered:current{model_name=~\"$model_name\"})", "legendFormat": "drift_features", "range": true, "refId": "A" } ], "title": "Critical Features in Drift", "type": "stat" }, { "datasource": { "type": "prometheus", "uid": "prometheus" }, "fieldConfig": { "defaults": { "color": { "mode": "palette-classic" }, "custom": { "axisCenteredZero": false, "drawStyle": "line", "fillOpacity": 10, "lineInterpolation": "linear", "lineWidth": 2, "showPoints": "never" }, "unit": "reqps" }, "overrides": [] }, "gridPos": { "h": 8, "w": 12, "x": 0, "y": 5 }, "id": 5, "options": { "legend": { "displayMode": "table", "placement": "bottom" }, "tooltip": { "mode": "multi" } }, "targets": [ { "editorMode": "code", "expr": "sum by (model_version) (job:ml_inference_rps:rate5m{model_name=~\"$model_name\",model_version=~\"$model_version\"})", "legendFormat": "{{ model_version }}", "range": true, "refId": "A" } ], "title": "Traffic by Model Version", "type": "timeseries" }, { "datasource": { "type": "prometheus", "uid": "prometheus" }, "fieldConfig": { "defaults": { "color": { "mode": "palette-classic" }, "custom": { "axisCenteredZero": false, "drawStyle": "line", "fillOpacity": 8, "lineInterpolation": "linear", "lineWidth": 2, "showPoints": "never" }, "unit": "s" }, "overrides": [] }, "gridPos": { "h": 8, "w": 12, "x": 12, "y": 5 }, "id": 6, "options": { "legend": { "displayMode": "table", "placement": "bottom" }, "tooltip": { "mode": "multi" } }, "targets": [ { "editorMode": "code", "expr": "job:ml_inference_latency_p95_seconds:5m{model_name=~\"$model_name\",model_version=~\"$model_version\"}", "legendFormat": "{{ model_version }} / {{ route }}", "range": true, "refId": "A" } ], "title": "P95 Latency by Version and Route", "type": "timeseries" }, { "datasource": { "type": "prometheus", "uid": "prometheus" }, "fieldConfig": { "defaults": { "color": { "mode": "palette-classic" }, "custom": { "axisCenteredZero": false, "drawStyle": "line", "fillOpacity": 8, "lineInterpolation": "linear", "lineWidth": 2, "showPoints": "never" } }, "overrides": [] }, "gridPos": { "h": 8, "w": 12, "x": 0, "y": 13 }, "id": 7, "options": { "legend": { "displayMode": "table", "placement": "bottom" }, "tooltip": { "mode": "multi" } }, "targets": [ { "editorMode": "code", "expr": "ml_data_drift_score{model_name=~\"$model_name\",feature_name=~\"$feature_name\"}", "legendFormat": "{{ feature_name }}", "range": true, "refId": "A" } ], "title": "Drift Score by Feature", "type": "timeseries" }, { "datasource": { "type": "prometheus", "uid": "prometheus" }, "fieldConfig": { "defaults": { "color": { "mode": "palette-classic" }, "custom": { "axisCenteredZero": false, "drawStyle": "line", "fillOpacity": 8, "lineInterpolation": "linear", "lineWidth": 2, "showPoints": "never" }, "unit": "percentunit" }, "overrides": [] }, "gridPos": { "h": 8, "w": 12, "x": 12, "y": 13 }, "id": 8, "options": { "legend": { "displayMode": "table", "placement": "bottom" }, "tooltip": { "mode": "multi" } }, "targets": [ { "editorMode": "code", "expr": "ml_feature_missing_ratio{model_name=~\"$model_name\",feature_name=~\"$feature_name\"}", "legendFormat": "{{ feature_name }}", "range": true, "refId": "A" } ], "title": "Missing Feature Ratio", "type": "timeseries" }, { "datasource": { "type": "prometheus", "uid": "prometheus" }, "fieldConfig": { "defaults": { "custom": { "align": "auto" } }, "overrides": [] }, "gridPos": { "h": 8, "w": 24, "x": 0, "y": 21 }, "id": 9, "options": { "cellHeight": "sm", "footer": { "show": false }, "showHeader": true }, "targets": [ { "editorMode": "code", "expr": "topk(10, ml_data_drift_score{model_name=~\"$model_name\"})", "format": "table", "instant": true, "legendFormat": "{{ feature_name }}", "refId": "A" } ], "title": "Top Drifted Features", "transformations": [ { "id": "organize", "options": { "excludeByName": { "__name__": true, "Time": true, "job": true, "instance": true }, "renameByName": { "Value": "Drift Score", "feature_name": "Feature", "model_name": "Model" } } } ], "type": "table" } ], "refresh": "30s", "schemaVersion": 39, "style": "dark", "tags": [ "ml", "monitoring", "prometheus", "grafana" ], "templating": { "list": [ { "current": { "selected": false, "text": ".*", "value": ".*" }, "datasource": { "type": "prometheus", "uid": "prometheus" }, "definition": "label_values(ml_model_info, model_name)", "hide": 0, "includeAll": false, "label": "Model", "multi": false, "name": "model_name", "options": [], "query": { "query": "label_values(ml_model_info, model_name)", "refId": "Prometheus-model_name-Variable-Query" }, "refresh": 1, "regex": "", "skipUrlSync": false, "sort": 1, "type": "query" }, { "current": { "selected": false, "text": ".*", "value": ".*" }, "datasource": { "type": "prometheus", "uid": "prometheus" }, "definition": "label_values(ml_model_info{model_name=~\"$model_name\"}, model_version)", "hide": 0, "includeAll": true, "label": "Model Version", "multi": true, "name": "model_version", "options": [], "query": { "query": "label_values(ml_model_info{model_name=~\"$model_name\"}, model_version)", "refId": "Prometheus-model_version-Variable-Query" }, "refresh": 1, "regex": "", "skipUrlSync": false, "sort": 1, "type": "query" }, { "current": { "selected": false, "text": ".*", "value": ".*" }, "datasource": { "type": "prometheus", "uid": "prometheus" }, "definition": "label_values(ml_data_drift_score{model_name=~\"$model_name\"}, feature_name)", "hide": 0, "includeAll": true, "label": "Feature", "multi": true, "name": "feature_name", "options": [], "query": { "query": "label_values(ml_data_drift_score{model_name=~\"$model_name\"}, feature_name)", "refId": "Prometheus-feature_name-Variable-Query" }, "refresh": 1, "regex": "", "skipUrlSync": false, "sort": 1, "type": "query" } ] }, "time": { "from": "now-6h", "to": "now" }, "timezone": "", "title": "ML Model Monitoring Overview", "uid": "ml-model-monitoring-overview", "version": 1, "weekStart": "" }