titan-iac/services/hermes/model-gate-deployment.yaml
2026-08-01 22:59:29 -03:00

126 lines
3.3 KiB
YAML

# services/hermes/model-gate-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: hermes-model-gate
namespace: hermes
labels:
app: hermes-model-gate
spec:
replicas: 1
revisionHistoryLimit: 2
selector:
matchLabels:
app: hermes-model-gate
template:
metadata:
labels:
app: hermes-model-gate
spec:
serviceAccountName: hermes-model-gate
securityContext:
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-04
- titan-13
- titan-15
- titan-17
- titan-19
preferredDuringSchedulingIgnoredDuringExecution:
- weight: 90
preference:
matchExpressions:
- key: hardware
operator: In
values:
- rpi5
containers:
- name: model-gate
image: python@sha256:6d43704baacd1bfbe7c295d7f13079d5d8104ed33568873133f8fc69980419df
imagePullPolicy: IfNotPresent
command:
- python
- /opt/model-gate/model_gate.py
ports:
- name: http
containerPort: 8080
env:
- name: UPSTREAM_URL
value: http://hermes-ollama.hermes.svc.cluster.local:11434
- name: LEASE_NAMESPACE
value: hermes
- name: LEASE_NAME
value: titan-24-gpu-owner
readinessProbe:
httpGet:
path: /healthz
port: http
initialDelaySeconds: 3
periodSeconds: 10
livenessProbe:
httpGet:
path: /healthz
port: http
initialDelaySeconds: 10
periodSeconds: 30
securityContext:
allowPrivilegeEscalation: false
capabilities:
drop:
- ALL
readOnlyRootFilesystem: true
runAsNonRoot: true
runAsUser: 65532
resources:
requests:
cpu: 25m
memory: 32Mi
limits:
cpu: 250m
memory: 128Mi
volumeMounts:
- name: script
mountPath: /opt/model-gate
readOnly: true
- name: tmp
mountPath: /tmp
volumes:
- name: script
configMap:
name: hermes-model-gate
defaultMode: 0555
- name: tmp
emptyDir: {}
---
apiVersion: v1
kind: Service
metadata:
name: hermes-model-gate
namespace: hermes
labels:
app: hermes-model-gate
spec:
type: ClusterIP
selector:
app: hermes-model-gate
ports:
- name: http
port: 11434
targetPort: http