feat: add zero-downtime deployment infrastructure

Adds PodDisruptionBudget, RollingUpdate strategy (maxUnavailable: 0),
preStop lifecycle hook, and startup probe support for zero-downtime deployments.

- PodDisruptionBudget: Ensures sufficient pods remain available during disruptions
- RollingUpdate maxUnavailable: 0: Prevents pod unavailability during updates
- preStop lifecycle hook: Allows graceful connection draining (10s delay)
- Startup probe: Gives slow-starting containers time to initialize (150s total)

Author: danieltruong

helm/eagle-api/templates/deployment.yaml

@@ -12,6 +12,11 @@ spec:
   {{- if not .Values.autoscaling.enabled }}
   replicas: {{ .Values.replicaCount }}
   {{- end }}
+  strategy:
+    type: RollingUpdate
+    rollingUpdate:
+      maxUnavailable: 0    # Never terminate pods until new ones are ready
+      maxSurge: 1          # Create one new pod at a time
   selector:
     matchLabels:
       {{- include "eagle-api.selectorLabels" . | nindent 6 }}
@@ -47,6 +52,18 @@ spec:
             initialDelaySeconds: {{ .Values.probes.readiness.initialDelaySeconds }}
             periodSeconds: {{ .Values.probes.readiness.periodSeconds }}
             timeoutSeconds: {{ .Values.probes.readiness.timeoutSeconds }}
+          {{- if .Values.probes.startup }}
+          startupProbe:
+            httpGet:
+              path: {{ .Values.probes.startup.path }}
+              port: http
+            failureThreshold: {{ .Values.probes.startup.failureThreshold }}
+            periodSeconds: {{ .Values.probes.startup.periodSeconds }}
+          {{- end }}
+          lifecycle:
+            preStop:
+              exec:
+                command: ["sh", "-c", "sleep 10"]  # Grace period for load balancer to drain
           resources:
             {{- toYaml .Values.resources | nindent 12 }}
           env:

helm/eagle-api/templates/pdb.yaml

@@ -0,0 +1,14 @@
+{{- if gt (int .Values.replicaCount) 1 }}
+apiVersion: policy/v1
+kind: PodDisruptionBudget
+metadata:
+  name: {{ include "eagle-api.fullname" . }}
+  namespace: {{ .Values.namespace }}
+  labels:
+    {{- include "eagle-api.labels" . | nindent 4 }}
+spec:
+  minAvailable: {{ .Values.podDisruptionBudget.minAvailable | default 2 }}
+  selector:
+    matchLabels:
+      {{- include "eagle-api.selectorLabels" . | nindent 6 }}
+{{- end }}

helm/eagle-api/values-prod.yaml

@@ -34,6 +34,10 @@ keycloak:
 persistence:
   size: 50Gi
 
+# Pod Disruption Budget - ensure at least 2 pods during disruptions
+podDisruptionBudget:
+  minAvailable: 2
+
 autoscaling:
   enabled: true
   minReplicas: 3

helm/eagle-api/values.yaml

@@ -47,6 +47,15 @@ probes:
     initialDelaySeconds: 10
     periodSeconds: 10
     timeoutSeconds: 10
+  # Startup probe for slow-starting containers (optional, enable in prod)
+  startup:
+    path: /api/docs
+    failureThreshold: 30
+    periodSeconds: 5
+
+# Pod Disruption Budget
+podDisruptionBudget:
+  minAvailable: 1  # Minimum pods available during disruptions
 
 # Persistent Volume Claim for document uploads
 persistence: