- Ensure prefix is passed for all services

- Add eveai-tem secret (Scaleway Transactional Email) to allow sending emails
- Adapted security URLs
- Certification problem in regions solved
- Redis insight added to tools in k8s
- Introduced new way of connection pooling for Redis
- TRA-79 - intrernal server error bij registreren catalog

scaleway/manifests/base/networking/README-ROUTING.md

@@ -1,14 +1,19 @@
 Routing alignment notes (staging/prod)
 
 Summary
-- Root (/) issues a 301 redirect to /admin/ via server-snippet on the apps ingress.
-- Prefixes /admin, /api, /chat-client are stripped at the edge and forwarded to their backends on /. The applications do not need to be prefix-aware.
+- Root (/) redirect will be handled at Bunny (edge). No server-snippet is used in the ingress.
+- Prefixes /admin, /api, /chat-client are stripped at the ingress and forwarded to their backends on /. The applications do not need to be prefix-routed internally.
+- For consistent external URLs (especially after POST/redirect), each prefix Ingress injects X-Forwarded-Prefix via nginx.ingress.kubernetes.io/proxy-set-headers and a per-prefix ConfigMap.
 - /verify remains available (Prefix) without any rewrite in a separate Ingress.
 - No CORS annotations at ingress. Static assets are served by Bunny CDN; API CORS is not handled here.
 - /flower is intentionally NOT exposed on k8s.
 
 Files
-- ingress-https.yaml: NGINX Ingress (apps) with regex paths and rewrite-target to strip prefixes; includes server-snippet to 301 redirect root to /admin/.
+- ingress-admin.yaml: Ingress for /admin with regex paths, rewrite-target to strip prefix, and proxy-set-headers pointing to eveai-admin-headers.
+- ingress-api.yaml: Ingress for /api with regex paths, rewrite-target to strip prefix, and proxy-set-headers pointing to eveai-api-headers.
+- ingress-chat-client.yaml: Ingress for /chat-client with regex paths, rewrite-target to strip prefix, and proxy-set-headers pointing to eveai-chat-headers.
+- headers-configmaps.yaml: ConfigMaps defining X-Forwarded-Prefix per ingress.
+- ingress-https.yaml: Legacy combined ingress retained for reference; server-snippet removed. Consider deprecating once split ingresses are applied.
 - ingress-verify.yaml: Separate Ingress for /verify without regex/rewrite.
 
 Paths behavior

scaleway/manifests/base/networking/headers-configmaps.yaml

@@ -0,0 +1,24 @@
+# ConfigMaps defining per-ingress proxy-set-headers
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: eveai-admin-headers
+  namespace: eveai-staging
+data:
+  X-Forwarded-Prefix: /admin
+---
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: eveai-api-headers
+  namespace: eveai-staging
+data:
+  X-Forwarded-Prefix: /api
+---
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: eveai-chat-headers
+  namespace: eveai-staging
+data:
+  X-Forwarded-Prefix: /chat-client

scaleway/manifests/base/networking/ingress-admin.yaml

@@ -0,0 +1,36 @@
+# Ingress for /admin prefix with X-Forwarded-Prefix header
+apiVersion: networking.k8s.io/v1
+kind: Ingress
+metadata:
+  name: eveai-admin-ingress
+  namespace: eveai-staging
+  labels:
+    app: eveai
+    environment: staging
+  annotations:
+    kubernetes.io/ingress.class: nginx
+    nginx.ingress.kubernetes.io/ssl-redirect: "true"
+    nginx.ingress.kubernetes.io/force-ssl-redirect: "true"
+    nginx.ingress.kubernetes.io/proxy-body-size: "10m"
+    nginx.ingress.kubernetes.io/proxy-read-timeout: "300"
+    cert-manager.io/cluster-issuer: letsencrypt-staging
+    nginx.ingress.kubernetes.io/use-regex: "true"
+    nginx.ingress.kubernetes.io/rewrite-target: "/$2"
+    nginx.ingress.kubernetes.io/proxy-set-headers: "eveai-staging/eveai-admin-headers"
+spec:
+  ingressClassName: nginx
+  tls:
+  - hosts:
+    - evie-staging.askeveai.com
+    secretName: evie-staging-tls
+  rules:
+  - host: evie-staging.askeveai.com
+    http:
+      paths:
+      - path: /admin(/|$)(.*)
+        pathType: ImplementationSpecific
+        backend:
+          service:
+            name: eveai-app-service
+            port:
+              number: 80

scaleway/manifests/base/networking/ingress-api.yaml

@@ -0,0 +1,36 @@
+# Ingress for /api prefix with X-Forwarded-Prefix header (optional but consistent)
+apiVersion: networking.k8s.io/v1
+kind: Ingress
+metadata:
+  name: eveai-api-ingress
+  namespace: eveai-staging
+  labels:
+    app: eveai
+    environment: staging
+  annotations:
+    kubernetes.io/ingress.class: nginx
+    nginx.ingress.kubernetes.io/ssl-redirect: "true"
+    nginx.ingress.kubernetes.io/force-ssl-redirect: "true"
+    nginx.ingress.kubernetes.io/proxy-body-size: "10m"
+    nginx.ingress.kubernetes.io/proxy-read-timeout: "300"
+    cert-manager.io/cluster-issuer: letsencrypt-staging
+    nginx.ingress.kubernetes.io/use-regex: "true"
+    nginx.ingress.kubernetes.io/rewrite-target: "/$2"
+    nginx.ingress.kubernetes.io/proxy-set-headers: "eveai-staging/eveai-api-headers"
+spec:
+  ingressClassName: nginx
+  tls:
+  - hosts:
+    - evie-staging.askeveai.com
+    secretName: evie-staging-tls
+  rules:
+  - host: evie-staging.askeveai.com
+    http:
+      paths:
+      - path: /api(/|$)(.*)
+        pathType: ImplementationSpecific
+        backend:
+          service:
+            name: eveai-api-service
+            port:
+              number: 80

scaleway/manifests/base/networking/ingress-chat-client.yaml

@@ -0,0 +1,36 @@
+# Ingress for /chat-client prefix with X-Forwarded-Prefix header (optional but consistent)
+apiVersion: networking.k8s.io/v1
+kind: Ingress
+metadata:
+  name: eveai-chat-client-ingress
+  namespace: eveai-staging
+  labels:
+    app: eveai
+    environment: staging
+  annotations:
+    kubernetes.io/ingress.class: nginx
+    nginx.ingress.kubernetes.io/ssl-redirect: "true"
+    nginx.ingress.kubernetes.io/force-ssl-redirect: "true"
+    nginx.ingress.kubernetes.io/proxy-body-size: "10m"
+    nginx.ingress.kubernetes.io/proxy-read-timeout: "300"
+    cert-manager.io/cluster-issuer: letsencrypt-staging
+    nginx.ingress.kubernetes.io/use-regex: "true"
+    nginx.ingress.kubernetes.io/rewrite-target: "/$2"
+    nginx.ingress.kubernetes.io/proxy-set-headers: "eveai-staging/eveai-chat-headers"
+spec:
+  ingressClassName: nginx
+  tls:
+  - hosts:
+    - evie-staging.askeveai.com
+    secretName: evie-staging-tls
+  rules:
+  - host: evie-staging.askeveai.com
+    http:
+      paths:
+      - path: /chat-client(/|$)(.*)
+        pathType: ImplementationSpecific
+        backend:
+          service:
+            name: eveai-chat-client-service
+            port:
+              number: 80

scaleway/manifests/base/networking/ingress-https.yaml

@@ -16,10 +16,6 @@ metadata:
     cert-manager.io/cluster-issuer: letsencrypt-staging
     nginx.ingress.kubernetes.io/use-regex: "true"
     nginx.ingress.kubernetes.io/rewrite-target: "/$2"
-    nginx.ingress.kubernetes.io/server-snippet: |
-      location = / {
-        return 301 /admin/;
-      }
 spec:
   ingressClassName: nginx
   tls:
@@ -30,28 +26,12 @@ spec:
   - host: evie-staging.askeveai.com
     http:
       paths:
-      # Application services (strip prefix)
-      - path: /admin(/|$)(.*)
-        pathType: ImplementationSpecific
+      - path: /verify
+        pathType: Prefix
         backend:
           service:
-            name: eveai-app-service
-            port:
-              number: 80
-      
-      - path: /api(/|$)(.*)
-        pathType: ImplementationSpecific
-        backend:
-          service:
-            name: eveai-api-service
-            port:
-              number: 80
-      
-      - path: /chat-client(/|$)(.*)
-        pathType: ImplementationSpecific
-        backend:
-          service:
-            name: eveai-chat-client-service
+            name: verify-service
             port:
               number: 80
+      # Application services (strip prefix) are now defined in dedicated ingress files