update dashboard and server load metric

fastmachinelearning · Jan 29, 2025 · d7818dd · d7818dd
1 parent 9d20e72
commit d7818dd
Show file tree

Hide file tree

Showing 2 changed files with 176 additions and 6 deletions.
diff --git a/helm/supersonic/dashboards/default.json b/helm/supersonic/dashboards/default.json
@@ -15,7 +15,7 @@
         "description": "",
         "gridPos": {
           "h": 4,
-          "w": 10,
+          "w": 24,
           "x": 0,
           "y": 0
         },
@@ -32,6 +32,77 @@
         "transparent": true,
         "type": "text"
     },
+    {
+        "datasource": {
+          "type": "prometheus",
+          "uid": "prometheus"
+        },
+        "fieldConfig": {
+          "defaults": {
+            "color": {
+              "mode": "thresholds"
+            },
+            "mappings": [],
+            "thresholds": {
+              "mode": "absolute",
+              "steps": [
+                {
+                  "color": "green",
+                  "value": null
+                },
+                {
+                  "color": "red",
+                  "value": 80
+                }
+              ]
+            }
+          },
+          "overrides": []
+        },
+        "gridPos": {
+          "h": 4,
+          "w": 4,
+          "x": 0,
+          "y": 4
+        },
+        "id": 4,
+        "options": {
+          "colorMode": "value",
+          "graphMode": "none",
+          "justifyMode": "auto",
+          "orientation": "auto",
+          "reduceOptions": {
+            "calcs": [
+              "lastNotNull"
+            ],
+            "fields": "",
+            "values": false
+          },
+          "textMode": "auto",
+          "wideLayout": true
+        },
+        "pluginVersion": "10.2.3",
+        "targets": [
+          {
+            "datasource": {
+              "type": "prometheus",
+              "uid": "prometheus"
+            },
+            "disableTextWrap": false,
+            "editorMode": "code",
+            "expr": "sum by(job)(envoy_cluster_membership_total{envoy_cluster_name=\"triton_grpc_service\", job=\"%RELEASE_NAME%\"})",
+            "fullMetaSearch": false,
+            "includeNullMetadata": true,
+            "instant": false,
+            "legendFormat": "{{job}}",
+            "range": true,
+            "refId": "A",
+            "useBackend": false
+          }
+        ],
+        "title": "Number of Triton Servers",
+        "type": "stat"
+    },
     {
       "datasource": {
         "type": "prometheus",
@@ -83,7 +154,7 @@
       "gridPos": {
         "h": 10,
         "w": 10,
-        "x": 0,
+        "x": 4,
         "y": 4
       },
       "id": 2,
@@ -112,6 +183,105 @@
       ],
       "title": "Server Load Metric",
       "type": "timeseries"
+    },
+    {
+        "datasource": {
+          "type": "prometheus",
+          "uid": "prometheus"
+        },
+        "fieldConfig": {
+          "defaults": {
+            "color": {
+              "mode": "palette-classic"
+            },
+            "custom": {
+              "axisBorderShow": false,
+              "axisCenteredZero": false,
+              "axisColorMode": "text",
+              "axisLabel": "",
+              "axisPlacement": "auto",
+              "barAlignment": 0,
+              "drawStyle": "line",
+              "fillOpacity": 0,
+              "gradientMode": "none",
+              "hideFrom": {
+                "legend": false,
+                "tooltip": false,
+                "viz": false
+              },
+              "insertNulls": false,
+              "lineInterpolation": "linear",
+              "lineWidth": 1,
+              "pointSize": 5,
+              "scaleDistribution": {
+                "type": "linear"
+              },
+              "showPoints": "auto",
+              "spanNulls": false,
+              "stacking": {
+                "group": "A",
+                "mode": "none"
+              },
+              "thresholdsStyle": {
+                "mode": "off"
+              }
+            },
+            "mappings": [],
+            "thresholds": {
+              "mode": "absolute",
+              "steps": [
+                {
+                  "color": "green",
+                  "value": null
+                },
+                {
+                  "color": "red",
+                  "value": 80
+                }
+              ]
+            }
+          },
+          "overrides": []
+        },
+        "gridPos": {
+          "h": 6,
+          "w": 4,
+          "x": 0,
+          "y": 8
+        },
+        "id": 3,
+        "options": {
+          "legend": {
+            "calcs": [],
+            "displayMode": "list",
+            "placement": "bottom",
+            "showLegend": true
+          },
+          "tooltip": {
+            "mode": "single",
+            "sort": "none"
+          }
+        },
+        "targets": [
+          {
+            "datasource": {
+              "type": "prometheus",
+              "uid": "prometheus"
+            },
+            "disableTextWrap": false,
+            "editorMode": "code",
+            "expr": "sum by(job)(envoy_cluster_membership_total{envoy_cluster_name=\"triton_grpc_service\", job=\"%RELEASE_NAME%\"})",
+            "fullMetaSearch": false,
+            "includeNullMetadata": true,
+            "instant": false,
+            "legendFormat": "{{job}}",
+            "range": true,
+            "refId": "A",
+            "useBackend": false
+          }
+        ],
+        "title": "Number of Triton Servers",
+        "type": "timeseries"
     }
   ],
   "refresh": "5s",

diff --git a/helm/supersonic/templates/_helpers.tpl b/helm/supersonic/templates/_helpers.tpl
@@ -27,12 +27,12 @@
 {{- if not ( eq .Values.prometheus.serverLoadMetric "" ) }}
   {{- printf "%s" .Values.prometheus.serverLoadMetric -}}
 {{- else }}
-sum by (instance) (
-    nv_inference_queue_duration_us{instance=~"{{ include "supersonic.name" . }}"}
+sum by (job) (
+    nv_inference_queue_duration_us{job=~"{{ include "supersonic.tritonName" . }}"}
 )
   /
-sum by (instance) (
-    (nv_inference_exec_count{instance=~"{{ include "supersonic.name" . }}"} * 1000) + 0.001
+sum by (job) (
+    (nv_inference_exec_count{job=~"{{ include "supersonic.tritonName" . }}"} * 1000) + 0.001
 )
 {{- end }}
 {{- end }}