From 65e99e2f6b270d3d4d74c958e490b8f9b09612d4 Mon Sep 17 00:00:00 2001
From: sd109 <scott@stackhpc.com>
Date: Thu, 4 Jul 2024 10:24:21 +0100
Subject: [PATCH] Pin remaining image tags for consistency

---
 .../templates/indexing-model-deployment.yaml  | 10 ++---
 .../templates/inference-model-deployment.yaml |  2 +-
 deployment/helm/values.yaml                   | 37 +++++++++++--------
 3 files changed, 27 insertions(+), 22 deletions(-)

diff --git a/deployment/helm/templates/indexing-model-deployment.yaml b/deployment/helm/templates/indexing-model-deployment.yaml
index 6986200235b..17a343b839a 100644
--- a/deployment/helm/templates/indexing-model-deployment.yaml
+++ b/deployment/helm/templates/indexing-model-deployment.yaml
@@ -7,7 +7,7 @@ metadata:
 spec:
   replicas: 1
   strategy:
-    {{- .Values.indexCapability.updateStrategy | toYaml | nindent 4 }}
+    {{- .Values.indexCapability.deployment.updateStrategy | toYaml | nindent 4 }}
   selector:
     matchLabels:
       {{- include "danswer-stack.selectorLabels" . | nindent 6 }}
@@ -28,12 +28,12 @@ spec:
     spec:
       containers:
       - name: indexing-model-server
-        image: danswer/danswer-model-server:latest
-        imagePullPolicy: IfNotPresent
+        image: "{{ .Values.indexCapability.deployment.image.repository }}:{{ .Values.indexCapability.deployment.image.tag | default .Chart.AppVersion }}"
+        imagePullPolicy: {{ .Values.indexCapability.deployment.image.pullPolicy }}
         command: [ "uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "9000", "--limit-concurrency", "10" ]
-        {{- if .Values.indexCapability.resources }}
+        {{- if .Values.indexCapability.deployment.resources }}
         resources:
-            {{- toYaml .Values.indexCapability.resources | nindent 10 }}
+            {{- toYaml .Values.indexCapability.deployment.resources | nindent 10 }}
         {{- end }}
         ports:
         - containerPort: 9000
diff --git a/deployment/helm/templates/inference-model-deployment.yaml b/deployment/helm/templates/inference-model-deployment.yaml
index 43caddd29c3..e91e2298e38 100644
--- a/deployment/helm/templates/inference-model-deployment.yaml
+++ b/deployment/helm/templates/inference-model-deployment.yaml
@@ -22,7 +22,7 @@ spec:
     spec:
       containers:
       - name: {{ .Values.inferenceCapability.service.name }}
-        image: {{ .Values.inferenceCapability.deployment.image.repository }}:{{ .Values.inferenceCapability.deployment.image.tag }}
+        image: "{{ .Values.inferenceCapability.deployment.image.repository }}:{{ .Values.inferenceCapability.deployment.image.tag | default .Chart.AppVersion }}"
         imagePullPolicy: {{ .Values.inferenceCapability.deployment.image.pullPolicy }}
         command: {{ toYaml .Values.inferenceCapability.deployment.command | nindent 14 }}
         ports:
diff --git a/deployment/helm/values.yaml b/deployment/helm/values.yaml
index ef358d454ca..272742ae975 100644
--- a/deployment/helm/values.yaml
+++ b/deployment/helm/values.yaml
@@ -24,7 +24,7 @@ inferenceCapability:
         value: inference-model-server
     image:
       repository: danswer/danswer-model-server
-      tag: latest
+      tag:
       pullPolicy: IfNotPresent
     command: ["uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "9000"]
     port: 9000
@@ -40,6 +40,23 @@ inferenceCapability:
       value: inference-model-server
 
 indexCapability:
+  deployment:
+    image:
+      repository: danswer/danswer-model-server
+      tag:
+      pullPolicy: IfNotPresent
+    resources:
+    # For example
+    # limits:
+    #   nvidia.com/gpu: 1
+    # The strategy to use for rolling out deployment updates
+    # If using GPU indexing with a limited number of GPUs available,
+    # this can be set to type: Recreate instead.
+    updateStrategy:
+      rollingUpdate:
+        maxSurge: 25%
+        maxUnavailable: 25%
+      type: RollingUpdate
   service:
     type: ClusterIP
     port: 9000
@@ -61,18 +78,6 @@ indexCapability:
     name: indexing-model-storage
     accessMode: "ReadWriteOnce"
     storage: "3Gi"
-  resources:
-    # For example
-    # limits:
-    #   nvidia.com/gpu: 1
-  # The strategy to use for rolling out deployment updates
-  # If using GPU indexing with a limited number of GPUs available,
-  # this can be set to type: Recreate instead.
-  updateStrategy:
-    rollingUpdate:
-      maxSurge: 25%
-      maxUnavailable: 25%
-    type: RollingUpdate
 
 
 config:
@@ -121,7 +126,7 @@ webserver:
     repository: danswer/danswer-web-server
     pullPolicy: IfNotPresent
     # Overrides the image tag whose default is the chart appVersion.
-    tag: ""
+    tag:
   deploymentLabels:
     app: web-server
   podAnnotations: {}
@@ -184,7 +189,7 @@ api:
     repository: danswer/danswer-backend
     pullPolicy: IfNotPresent
     # Overrides the image tag whose default is the chart appVersion.
-    tag: ""
+    tag:
   deploymentLabels:
     app: api-server
   podAnnotations: {}
@@ -249,7 +254,7 @@ background:
     repository: danswer/danswer-backend
     pullPolicy: IfNotPresent
     # Overrides the image tag whose default is the chart appVersion.
-    tag: latest
+    tag:
   podAnnotations: {}
   podLabels:
     scope: danswer-backend