feat: add Redis Cluster mode support (GCP Memorystore, AWS ElastiCache)

helm/codeapi/README.md

@@ -58,16 +58,16 @@ verifier through environment variables on the api component, e.g.:
 
 ```yaml
 api:
-  extraEnv:
-    - name: CODEAPI_AUTH_PROVIDER
-      value: librechat-jwt
-    - name: CODEAPI_JWT_PUBLIC_KEY     # single PEM/base64-DER verifier key
-      valueFrom:
-        secretKeyRef:
-          name: codeapi-jwt-verifier
-          key: public-key
-    - name: CODEAPI_JWT_KID
-      value: my-key-id
+    extraEnv:
+        - name: CODEAPI_AUTH_PROVIDER
+          value: librechat-jwt
+        - name: CODEAPI_JWT_PUBLIC_KEY # single PEM/base64-DER verifier key
+          valueFrom:
+              secretKeyRef:
+                  name: codeapi-jwt-verifier
+                  key: public-key
+        - name: CODEAPI_JWT_KID
+          value: my-key-id
 ```
 
 `CODEAPI_JWT_PUBLIC_KEYS_DIR` (a mounted directory of PEM files) and
@@ -78,14 +78,35 @@ For development only, `LOCAL_MODE=true` bypasses authentication — see
 **TLS to Redis.** Set `REDIS_TLS=true` via `extraEnv` on each component when
 your Redis (e.g. a managed cache) requires TLS.
 
+**Redis Cluster mode (GCP Memorystore cluster, AWS ElastiCache cluster).** Set
+`redis.cluster.enabled=true` and provide the comma-separated startup nodes in
+`redis.cluster.nodes`. The chart will set `USE_REDIS_CLUSTER=true` and populate
+`REDIS_HOST` with the node list on every component. For TLS with a CA certificate
+(recommended for GCP Memorystore), also set:
+
+```yaml
+redis:
+    tls:
+        enabled: true
+        caSecretName: my-memorystore-secret # Secret that holds the CA cert
+        caKey: ca # Key inside the Secret
+        caMountPath: /etc/redis-tls/ca.crt # Mount path in each pod
+    useAlternativeDnsLookup: true # Required for GCP Memorystore cluster
+```
+
+The CA certificate is mounted from the Secret into every component pod and
+passed to ioredis via `REDIS_CA`. No sidecar is needed.
+
 ## Quick Start (Local Development)
 
 ### 1. Start Minikube
+
 ```bash
 minikube start --cpus=4 --memory=8192
 ```
 
 ### 2. Build Images Inside Minikube
+
 ```bash
 # Point docker to minikube's daemon
 eval $(minikube docker-env)
@@ -100,6 +121,7 @@ docker build -t codeapi-package-init:latest -f docker/Dockerfile.package-init .
 ```
 
 ### 3. Install Dependencies & Deploy
+
 ```bash
 cd helm/codeapi
 
@@ -151,6 +173,7 @@ kubectl rollout restart deployment/codeapi-sandbox-runner
 ```
 
 ### 5. Access the API
+
 ```bash
 # Port forward (in another terminal)
 kubectl port-forward svc/codeapi-api 3112:3112
@@ -164,6 +187,7 @@ curl http://localhost:3112/v1/health
 ## Commands Reference
 
 ### Startup
+
 ```bash
 # Start minikube
 minikube start
@@ -176,6 +200,7 @@ kubectl port-forward svc/codeapi-api 3112:3112
 ```
 
 ### Check Status
+
 ```bash
 # View all pods
 kubectl get pods
@@ -190,6 +215,7 @@ kubectl describe pod <pod-name>
 ```
 
 ### Scaling
+
 ```bash
 # Scale the sandbox execution tier
 kubectl scale deployment/codeapi-sandbox-runner --replicas=10
@@ -200,6 +226,7 @@ helm upgrade codeapi ./helm/codeapi -f ./helm/codeapi/values-local.yaml \
 ```
 
 ### Update After Code Changes
+
 ```bash
 # Rebuild images (must be in minikube docker env)
 eval $(minikube docker-env)
@@ -212,6 +239,7 @@ kubectl rollout restart deployment/codeapi-sandbox-runner
 ```
 
 ### Teardown
+
 ```bash
 # Uninstall the Helm release (removes all K8s resources)
 helm uninstall codeapi
@@ -228,12 +256,14 @@ minikube delete
 ## Testing
 
 ### Health Check
+
 ```bash
 curl http://localhost:3112/v1/health
 # Expected: OK
 ```
 
 ### Execute Python Code
+
 ```bash
 curl -X POST http://localhost:3112/v1/exec \
   -H "Content-Type: application/json" \
@@ -242,6 +272,7 @@ curl -X POST http://localhost:3112/v1/exec \
 ```
 
 ### Verify Horizontal Scaling
+
 ```bash
 # Check which service-worker processed the job
 kubectl logs deployment/codeapi-service-worker --tail=5
@@ -299,6 +330,7 @@ kubectl logs deployment/codeapi-service-worker --tail=5
 ## Troubleshooting
 
 ### Pod stuck in `ErrImageNeverPull`
+
 ```bash
 # Images must be built inside minikube's docker
 eval $(minikube docker-env)
@@ -307,13 +339,15 @@ kubectl rollout restart deployment/<deployment-name>
 ```
 
 ### Pod stuck in `CrashLoopBackOff`
+
 ```bash
 # Check logs
 kubectl logs <pod-name> --previous
 kubectl describe pod <pod-name>
 ```
 
 ### "runtime is unknown" error
+
 ```bash
 # Language packages PVC is empty. Check if the init job ran:
 kubectl get jobs -l app.kubernetes.io/component=package-init
@@ -327,12 +361,14 @@ kubectl rollout restart deployment/codeapi-sandbox-runner
 ```
 
 ### Connection refused on port 3112
+
 ```bash
 # Make sure port-forward is running
 kubectl port-forward svc/codeapi-api 3112:3112
 ```
 
 ### MinIO `ImagePullBackOff` (production values)
+
 ```bash
 # The Bitnami MinIO chart may reference unavailable image tags.
 # For local dev, values-local.yaml uses minio.useSimple=true which

helm/codeapi/templates/_helpers.tpl

@@ -152,11 +152,15 @@ app.kubernetes.io/component: tool-call-server
 {{- end }}
 
 {{/*
-Redis host - either from subchart or external
+Redis host - either from subchart or external.
+In cluster mode this renders the comma-separated node list from redis.cluster.nodes
+(falling back to redis.external.host for single-node external deployments).
 */}}
 {{- define "codeapi.redis.host" -}}
 {{- if .Values.redis.enabled }}
 {{- printf "%s-redis-master" .Release.Name }}
+{{- else if and .Values.redis.cluster.enabled .Values.redis.cluster.nodes }}
+{{- .Values.redis.cluster.nodes }}
 {{- else }}
 {{- .Values.redis.external.host }}
 {{- end }}
@@ -173,6 +177,70 @@ Redis port
 {{- end }}
 {{- end }}
 
+{{/*
+USE_REDIS_CLUSTER value – "true" when redis.cluster.enabled or when
+redis.cluster.nodes contains a comma (auto-detect multiple nodes).
+*/}}
+{{- define "codeapi.redis.clusterEnabled" -}}
+{{- if .Values.redis.cluster.enabled }}
+{{- "true" }}
+{{- else if and .Values.redis.cluster.nodes (contains "," .Values.redis.cluster.nodes) }}
+{{- "true" }}
+{{- else }}
+{{- "false" }}
+{{- end }}
+{{- end }}
+
+{{/*
+Emit the Redis TLS + CA environment variables and volume mount for each
+component that needs it.  Renders nothing when redis.tls.enabled is false.
+Usage: {{ include "codeapi.redis.tlsEnv" . }}
+*/}}
+{{- define "codeapi.redis.tlsEnv" -}}
+{{- if .Values.redis.tls.enabled }}
+- name: REDIS_TLS
+  value: "true"
+{{- if .Values.redis.tls.caSecretName }}
+- name: REDIS_CA
+  value: {{ .Values.redis.tls.caMountPath | quote }}
+{{- end }}
+{{- end }}
+{{- if .Values.redis.useAlternativeDnsLookup }}
+- name: REDIS_USE_ALTERNATIVE_DNS_LOOKUP
+  value: "true"
+{{- end }}
+{{- end }}
+
+{{/*
+Volume definition for the Redis CA certificate secret.
+Renders nothing when redis.tls.caSecretName is empty.
+Usage: {{ include "codeapi.redis.caVolume" . }}
+*/}}
+{{- define "codeapi.redis.caVolume" -}}
+{{- if and .Values.redis.tls.enabled .Values.redis.tls.caSecretName }}
+- name: redis-ca
+  secret:
+    secretName: {{ .Values.redis.tls.caSecretName }}
+    items:
+      - key: {{ .Values.redis.tls.caKey }}
+        path: ca.crt
+{{- end }}
+{{- end }}
+
+{{/*
+VolumeMount for the Redis CA certificate inside a container.
+Renders nothing when redis.tls.caSecretName is empty.
+Usage: {{ include "codeapi.redis.caVolumeMount" . }}
+*/}}
+{{- define "codeapi.redis.caVolumeMount" -}}
+{{- if and .Values.redis.tls.enabled .Values.redis.tls.caSecretName }}
+- name: redis-ca
+  mountPath: {{ .Values.redis.tls.caMountPath | quote }}
+  subPath: ca.crt
+  readOnly: true
+{{- end }}
+{{- end }}
+
 {{/*
 Redis NetworkPolicy egress. Kubernetes NetworkPolicy cannot match DNS names,
 so external Redis can be scoped with CIDRs when available; otherwise the chart

helm/codeapi/templates/api-deployment.yaml

@@ -51,7 +51,9 @@ spec:
                 secretKeyRef:
                   name: {{ include "codeapi.fullname" . }}-secrets
                   key: redis-password
-            # Service URLs
+            - name: USE_REDIS_CLUSTER
+              value: {{ include "codeapi.redis.clusterEnabled" . | quote }}
+            {{- include "codeapi.redis.tlsEnv" . | nindent 12 }}
             - name: FILE_SERVER_URL
               value: "http://{{ include "codeapi.fullname" . }}-file-server:{{ .Values.fileServer.service.port }}"
             - name: TOOL_CALL_SERVER_URL
@@ -87,6 +89,14 @@ spec:
             periodSeconds: 10
           resources:
             {{- toYaml .Values.api.resources | nindent 12 }}
+          {{- with (include "codeapi.redis.caVolumeMount" . | trim) }}
+          volumeMounts:
+            {{- . | nindent 12 }}
+          {{- end }}
+      {{- with (include "codeapi.redis.caVolume" . | trim) }}
+      volumes:
+        {{- . | nindent 8 }}
+      {{- end }}
       {{- with .Values.nodeSelector }}
       nodeSelector:
         {{- toYaml . | nindent 8 }}

helm/codeapi/templates/egress-gateway-deployment.yaml

@@ -69,6 +69,9 @@ spec:
                 secretKeyRef:
                   name: {{ include "codeapi.fullname" . }}-secrets
                   key: redis-password
+            - name: USE_REDIS_CLUSTER
+              value: {{ include "codeapi.redis.clusterEnabled" . | quote }}
+            {{- include "codeapi.redis.tlsEnv" . | nindent 12 }}
           livenessProbe:
             httpGet:
               path: /live
@@ -83,6 +86,14 @@ spec:
             periodSeconds: 10
           resources:
             {{- toYaml .Values.egressGateway.resources | nindent 12 }}
+          {{- with (include "codeapi.redis.caVolumeMount" . | trim) }}
+          volumeMounts:
+            {{- . | nindent 12 }}
+          {{- end }}
+      {{- with (include "codeapi.redis.caVolume" . | trim) }}
+      volumes:
+        {{- . | nindent 8 }}
+      {{- end }}
       {{- with .Values.nodeSelector }}
       nodeSelector:
         {{- toYaml . | nindent 8 }}

helm/codeapi/templates/file-server-deployment.yaml

@@ -110,6 +110,9 @@ spec:
                 secretKeyRef:
                   name: {{ include "codeapi.fullname" . }}-secrets
                   key: redis-password
+            - name: USE_REDIS_CLUSTER
+              value: {{ include "codeapi.redis.clusterEnabled" . | quote }}
+            {{- include "codeapi.redis.tlsEnv" . | nindent 12 }}
             - name: CODEAPI_INTERNAL_SERVICE_TOKEN
               valueFrom:
                 secretKeyRef:
@@ -135,6 +138,14 @@ spec:
             timeoutSeconds: 5
           resources:
             {{- toYaml .Values.fileServer.resources | nindent 12 }}
+          {{- with (include "codeapi.redis.caVolumeMount" . | trim) }}
+          volumeMounts:
+            {{- . | nindent 12 }}
+          {{- end }}
+      {{- with (include "codeapi.redis.caVolume" . | trim) }}
+      volumes:
+        {{- . | nindent 8 }}
+      {{- end }}
       {{- with .Values.nodeSelector }}
       nodeSelector:
         {{- toYaml . | nindent 8 }}

helm/codeapi/templates/tool-call-server-deployment.yaml

@@ -44,6 +44,9 @@ spec:
                 secretKeyRef:
                   name: {{ include "codeapi.fullname" . }}-secrets
                   key: redis-password
+            - name: USE_REDIS_CLUSTER
+              value: {{ include "codeapi.redis.clusterEnabled" . | quote }}
+            {{- include "codeapi.redis.tlsEnv" . | nindent 12 }}
             - name: CODEAPI_INTERNAL_SERVICE_TOKEN
               valueFrom:
                 secretKeyRef:
@@ -70,6 +73,14 @@ spec:
             periodSeconds: 10
           resources:
             {{- toYaml .Values.toolCallServer.resources | nindent 12 }}
+          {{- with (include "codeapi.redis.caVolumeMount" . | trim) }}
+          volumeMounts:
+            {{- . | nindent 12 }}
+          {{- end }}
+      {{- with (include "codeapi.redis.caVolume" . | trim) }}
+      volumes:
+        {{- . | nindent 8 }}
+      {{- end }}
       {{- with .Values.nodeSelector }}
       nodeSelector:
         {{- toYaml . | nindent 8 }}