6183159974
- Install Prometheus + Grafana via kube-prometheus-stack (ClusterIP only, no public ingress) - Deploy Tailscale Operator for secure VPN access to internal services - Add CNPG/PostgreSQL monitoring dashboards - Add vLLM inference monitoring dashboards (tokens, latency, GPU) - Add Cilium networking dashboards (policy, traffic, drops) - Update infra-controllers staging kustomization to include all controllers - Add monitoring-configs Flux sync for dashboard deployment - Update README with monitoring architecture and access instructions - Remove broken stale monitoring files (Azure Key Vault refs, wrong domains) Access: kubectl port-forward or Tailscale VPN (replace auth key before deploy)
70 lines
1.8 KiB
YAML
70 lines
1.8 KiB
YAML
apiVersion: apps/v1
|
|
kind: Deployment
|
|
metadata:
|
|
name: vwap-monitor
|
|
namespace: customer1
|
|
labels:
|
|
app: vwap-monitor
|
|
component: scanner
|
|
spec:
|
|
replicas: 1
|
|
strategy:
|
|
type: Recreate # only one instance should run
|
|
selector:
|
|
matchLabels:
|
|
app: vwap-monitor
|
|
template:
|
|
metadata:
|
|
labels:
|
|
app: vwap-monitor
|
|
component: scanner
|
|
annotations:
|
|
# Restart if config changes
|
|
checksum/config: "vwap-monitor-config"
|
|
spec:
|
|
terminationGracePeriodSeconds: 30
|
|
containers:
|
|
- name: scanner
|
|
image: us-central1-docker.pkg.dev/devops-lab-cluster/customer1/vwap-monitor:latest
|
|
imagePullPolicy: Always
|
|
resources:
|
|
requests:
|
|
cpu: 100m
|
|
memory: 64Mi
|
|
limits:
|
|
cpu: 250m
|
|
memory: 256Mi
|
|
envFrom:
|
|
- configMapRef:
|
|
name: vwap-monitor-config
|
|
env:
|
|
- name: TELEGRAM_BOT_TOKEN
|
|
valueFrom:
|
|
secretKeyRef:
|
|
name: vwap-monitor-secrets
|
|
key: telegram-bot-token
|
|
- name: TELEGRAM_CHAT_ID
|
|
valueFrom:
|
|
secretKeyRef:
|
|
name: vwap-monitor-secrets
|
|
key: telegram-chat-id
|
|
startupProbe:
|
|
exec:
|
|
command: ["/bin/sh", "-c", "python3 -c 'import scanner'"]
|
|
initialDelaySeconds: 10
|
|
periodSeconds: 10
|
|
failureThreshold: 3
|
|
livenessProbe:
|
|
exec:
|
|
command: ["/bin/sh", "-c", "kill -0 1"]
|
|
initialDelaySeconds: 30
|
|
periodSeconds: 60
|
|
---
|
|
apiVersion: v1
|
|
kind: ConfigMap
|
|
metadata:
|
|
name: vwap-monitor-config
|
|
namespace: customer1
|
|
data:
|
|
SCAN_INTERVAL_SEC: "120"
|
|
BREACH_THRESHOLD: "2.0"
|