titan-iac/services/hermes/switchyard-deployment.yaml

# services/hermes/switchyard-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: hermes-switchyard
  namespace: hermes
  labels:
    app: hermes-switchyard
spec:
  replicas: 1
  revisionHistoryLimit: 2
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 0
  selector:
    matchLabels:
      app: hermes-switchyard
  template:
    metadata:
      labels:
        app: hermes-switchyard
      annotations:
        ai.bstein.dev/config-rev: "20260812-image-continuity"
        prometheus.io/scrape: "true"
        prometheus.io/port: "9005"
        prometheus.io/path: /metrics
        vault.hashicorp.com/agent-inject: "true"
        vault.hashicorp.com/role: hermes-agent
        vault.hashicorp.com/agent-inject-secret-anthropic-token: kv/data/atlas/hermes/agent-tokens
        vault.hashicorp.com/agent-inject-template-anthropic-token: |
          {{- with secret "kv/data/atlas/hermes/agent-tokens" -}}
          {{ .Data.data.anthropic_oauth_token }}
          {{- end }}
        vault.hashicorp.com/agent-inject-secret-relay-key: kv/data/atlas/hermes/chat-telegram
        vault.hashicorp.com/agent-inject-template-relay-key: |
          {{- with secret "kv/data/atlas/hermes/chat-telegram" -}}
          {{ .Data.data.relay_key }}
          {{- end }}
        vault.hashicorp.com/agent-pre-populate-only: "true"
        vault.hashicorp.com/agent-init-first: "true"
        vault.hashicorp.com/agent-requests-cpu: 25m
        vault.hashicorp.com/agent-requests-mem: 32Mi
        vault.hashicorp.com/agent-limits-cpu: 100m
        vault.hashicorp.com/agent-limits-mem: 128Mi
    spec:
      serviceAccountName: hermes-switchyard
      automountServiceAccountToken: true
      terminationGracePeriodSeconds: 330
      securityContext:
        fsGroup: 10000
        fsGroupChangePolicy: OnRootMismatch
        seccompProfile:
          type: RuntimeDefault
      affinity:
        nodeAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
            nodeSelectorTerms:
              - matchExpressions:
                  - key: kubernetes.io/arch
                    operator: In
                    values: [arm64]
                  - key: node-role.kubernetes.io/worker
                    operator: In
                    values: ["true"]
                  - key: kubernetes.io/hostname
                    operator: NotIn
                    values: [titan-14, titan-18, titan-22, titan-24]
      containers:
        - name: switchyard
          image: registry.bstein.dev/bstein/hermes-switchyard@sha256:3d952d528a4e4cb8afdf84f8995272f5d6107dc127b68cc20234b1da4aff43eb
          imagePullPolicy: IfNotPresent
          command: [/bin/sh, -ec]
          args:
            - |
              ATLAS_BROKER_KEY="$(tr -d '\r\n' < /vault/secrets/relay-key)"
              export ATLAS_BROKER_KEY
              exec switchyard-server \
                --config /etc/switchyard/routes.toml \
                --host 0.0.0.0 \
                --port 9005 \
                --shutdown-timeout 5m \
                --routing-log-file /var/lib/switchyard/routing.jsonl
          env:
            - name: RUST_LOG
              value: switchyard_server=info,libsy=info
          ports:
            - name: http
              containerPort: 9005
              protocol: TCP
          readinessProbe:
            httpGet:
              path: /health
              port: http
            initialDelaySeconds: 3
            periodSeconds: 10
            timeoutSeconds: 3
          livenessProbe:
            httpGet:
              path: /health
              port: http
            initialDelaySeconds: 15
            periodSeconds: 30
            timeoutSeconds: 5
          securityContext:
            allowPrivilegeEscalation: false
            capabilities:
              drop: [ALL]
            readOnlyRootFilesystem: true
            runAsNonRoot: true
            runAsUser: 10000
            runAsGroup: 10000
          resources:
            requests:
              cpu: 100m
              memory: 128Mi
            limits:
              cpu: "2"
              memory: 1Gi
          volumeMounts:
            - name: config
              mountPath: /etc/switchyard
              readOnly: true
            - name: state
              mountPath: /var/lib/switchyard
            - name: tmp
              mountPath: /tmp
        - name: claude-oauth-broker
          image: registry.bstein.dev/bstein/hermes-switchyard-brokers@sha256:ee7e95e060ef8083da505162d7e9030daba15fdd828cc047bbcbe6aa409d2083
          imagePullPolicy: IfNotPresent
          command:
            - python
            - /opt/coordinator/claude_oauth_broker.py
          env:
            - name: HERMES_CLAUDE_OAUTH_TOKEN_FILE
              value: /vault/secrets/anthropic-token
            - name: HERMES_CLAUDE_BROKER_KEY_FILE
              value: /vault/secrets/relay-key
            - name: HERMES_CLAUDE_BROKER_READ_TIMEOUT
              value: "1800"
            - name: HERMES_ROUTING_CATALOG_PATH
              value: /routing-catalog/catalog.json
          ports:
            - name: claude
              containerPort: 9006
              protocol: TCP
          readinessProbe:
            httpGet:
              path: /health
              port: claude
            initialDelaySeconds: 3
            periodSeconds: 10
            timeoutSeconds: 3
          livenessProbe:
            httpGet:
              path: /health
              port: claude
            initialDelaySeconds: 15
            periodSeconds: 30
            timeoutSeconds: 5
          securityContext:
            allowPrivilegeEscalation: false
            capabilities:
              drop: [ALL]
            readOnlyRootFilesystem: true
            runAsNonRoot: true
            runAsUser: 10000
            runAsGroup: 10000
          resources:
            requests:
              cpu: 50m
              memory: 96Mi
            limits:
              cpu: 500m
              memory: 384Mi
          volumeMounts:
            - name: coordinator
              mountPath: /opt/coordinator
              readOnly: true
            - name: tmp
              mountPath: /tmp
            - name: routing-catalog
              mountPath: /routing-catalog
              readOnly: true
        - name: worker-route-broker
          image: registry.bstein.dev/bstein/hermes-switchyard-brokers@sha256:ee7e95e060ef8083da505162d7e9030daba15fdd828cc047bbcbe6aa409d2083
          imagePullPolicy: IfNotPresent
          command:
            - python
            - /opt/coordinator/worker_route_broker.py
          ports:
            - name: worker-route
              containerPort: 9007
              protocol: TCP
          env:
            - name: HERMES_ROUTING_CATALOG_PATH
              value: /routing-catalog/catalog.json
          readinessProbe:
            httpGet:
              path: /health
              port: worker-route
            initialDelaySeconds: 2
            periodSeconds: 10
            timeoutSeconds: 3
          livenessProbe:
            httpGet:
              path: /health
              port: worker-route
            initialDelaySeconds: 10
            periodSeconds: 30
            timeoutSeconds: 5
          securityContext:
            allowPrivilegeEscalation: false
            capabilities:
              drop: [ALL]
            readOnlyRootFilesystem: true
            runAsNonRoot: true
            runAsUser: 10000
            runAsGroup: 10000
          resources:
            requests:
              cpu: 10m
              memory: 24Mi
            limits:
              cpu: 100m
              memory: 96Mi
          volumeMounts:
            - name: coordinator
              mountPath: /opt/coordinator
              readOnly: true
            - name: tmp
              mountPath: /tmp
            - name: routing-catalog
              mountPath: /routing-catalog
              readOnly: true
      volumes:
        - name: config
          configMap:
            name: hermes-switchyard
        - name: coordinator
          configMap:
            name: hermes-coordinator
            defaultMode: 0555
        - name: state
          persistentVolumeClaim:
            claimName: hermes-switchyard-state-rwx
        - name: routing-catalog
          persistentVolumeClaim:
            claimName: hermes-routing-catalog
        - name: tmp
          emptyDir: {}