Add role-based filtering and imporve code
This commit is contained in:
Urtzi Alfaro
@@ -19,14 +19,14 @@ spec:
|
||||
spec:
|
||||
serviceAccountName: demo-session-sa
|
||||
containers:
|
||||
- name: demo-session
|
||||
- name: demo-session-service
|
||||
image: bakery/demo-session-service:latest
|
||||
ports:
|
||||
- containerPort: 8000
|
||||
name: http
|
||||
env:
|
||||
- name: SERVICE_NAME
|
||||
value: "demo-session"
|
||||
value: "demo-session-service"
|
||||
- name: DEMO_SESSION_DATABASE_URL
|
||||
valueFrom:
|
||||
secretKeyRef:
|
||||
@@ -82,3 +82,14 @@ spec:
|
||||
port: 8000
|
||||
initialDelaySeconds: 10
|
||||
periodSeconds: 10
|
||||
startupProbe:
|
||||
httpGet:
|
||||
path: /health
|
||||
port: 8000
|
||||
initialDelaySeconds: 10
|
||||
periodSeconds: 5
|
||||
failureThreshold: 30
|
||||
initContainers:
|
||||
- name: wait-for-redis
|
||||
image: busybox:1.36
|
||||
command: ['sh', '-c', 'until nc -z redis-service 6379; do echo waiting for redis; sleep 2; done']
|
||||
|
||||
@@ -0,0 +1,177 @@
|
||||
apiVersion: v1
|
||||
kind: ConfigMap
|
||||
metadata:
|
||||
name: grafana-dashboards
|
||||
namespace: monitoring
|
||||
data:
|
||||
gateway-metrics.json: |
|
||||
{
|
||||
"dashboard": {
|
||||
"title": "Bakery IA - Gateway Metrics",
|
||||
"tags": ["bakery-ia", "gateway"],
|
||||
"timezone": "browser",
|
||||
"panels": [
|
||||
{
|
||||
"id": 1,
|
||||
"title": "Request Rate by Endpoint",
|
||||
"type": "graph",
|
||||
"gridPos": {"x": 0, "y": 0, "w": 12, "h": 8},
|
||||
"targets": [{
|
||||
"expr": "rate(http_requests_total{service=\"gateway\"}[5m])",
|
||||
"legendFormat": "{{method}} {{endpoint}}"
|
||||
}]
|
||||
},
|
||||
{
|
||||
"id": 2,
|
||||
"title": "P95 Request Latency",
|
||||
"type": "graph",
|
||||
"gridPos": {"x": 12, "y": 0, "w": 12, "h": 8},
|
||||
"targets": [{
|
||||
"expr": "histogram_quantile(0.95, rate(http_request_duration_seconds_bucket{service=\"gateway\"}[5m]))",
|
||||
"legendFormat": "{{endpoint}} p95"
|
||||
}]
|
||||
},
|
||||
{
|
||||
"id": 3,
|
||||
"title": "Error Rate (5xx)",
|
||||
"type": "graph",
|
||||
"gridPos": {"x": 0, "y": 8, "w": 12, "h": 8},
|
||||
"targets": [{
|
||||
"expr": "rate(http_requests_total{service=\"gateway\",status_code=~\"5..\"}[5m])",
|
||||
"legendFormat": "{{endpoint}} errors"
|
||||
}]
|
||||
},
|
||||
{
|
||||
"id": 4,
|
||||
"title": "Active Requests",
|
||||
"type": "stat",
|
||||
"gridPos": {"x": 12, "y": 8, "w": 6, "h": 4},
|
||||
"targets": [{
|
||||
"expr": "sum(rate(http_requests_total{service=\"gateway\"}[1m]))"
|
||||
}]
|
||||
},
|
||||
{
|
||||
"id": 5,
|
||||
"title": "Authentication Success Rate",
|
||||
"type": "stat",
|
||||
"gridPos": {"x": 18, "y": 8, "w": 6, "h": 4},
|
||||
"targets": [{
|
||||
"expr": "rate(gateway_auth_responses_total[5m]) / rate(gateway_auth_requests_total[5m]) * 100"
|
||||
}]
|
||||
}
|
||||
],
|
||||
"refresh": "10s",
|
||||
"schemaVersion": 16,
|
||||
"version": 1
|
||||
}
|
||||
}
|
||||
|
||||
services-overview.json: |
|
||||
{
|
||||
"dashboard": {
|
||||
"title": "Bakery IA - Services Overview",
|
||||
"tags": ["bakery-ia", "services"],
|
||||
"timezone": "browser",
|
||||
"panels": [
|
||||
{
|
||||
"id": 1,
|
||||
"title": "Request Rate by Service",
|
||||
"type": "graph",
|
||||
"gridPos": {"x": 0, "y": 0, "w": 12, "h": 8},
|
||||
"targets": [{
|
||||
"expr": "sum by (service) (rate(http_requests_total[5m]))",
|
||||
"legendFormat": "{{service}}"
|
||||
}]
|
||||
},
|
||||
{
|
||||
"id": 2,
|
||||
"title": "P99 Latency by Service",
|
||||
"type": "graph",
|
||||
"gridPos": {"x": 12, "y": 0, "w": 12, "h": 8},
|
||||
"targets": [{
|
||||
"expr": "histogram_quantile(0.99, sum by (service, le) (rate(http_request_duration_seconds_bucket[5m])))",
|
||||
"legendFormat": "{{service}} p99"
|
||||
}]
|
||||
},
|
||||
{
|
||||
"id": 3,
|
||||
"title": "Error Rate by Service",
|
||||
"type": "graph",
|
||||
"gridPos": {"x": 0, "y": 8, "w": 24, "h": 8},
|
||||
"targets": [{
|
||||
"expr": "sum by (service) (rate(http_requests_total{status_code=~\"5..\"}[5m]))",
|
||||
"legendFormat": "{{service}}"
|
||||
}]
|
||||
},
|
||||
{
|
||||
"id": 4,
|
||||
"title": "Service Health Status",
|
||||
"type": "table",
|
||||
"gridPos": {"x": 0, "y": 16, "w": 24, "h": 8},
|
||||
"targets": [{
|
||||
"expr": "up{job=\"bakery-services\"}",
|
||||
"format": "table",
|
||||
"instant": true
|
||||
}],
|
||||
"transformations": [{
|
||||
"id": "organize",
|
||||
"options": {
|
||||
"excludeByName": {},
|
||||
"indexByName": {},
|
||||
"renameByName": {
|
||||
"service": "Service Name",
|
||||
"Value": "Status"
|
||||
}
|
||||
}
|
||||
}]
|
||||
}
|
||||
],
|
||||
"refresh": "30s",
|
||||
"schemaVersion": 16,
|
||||
"version": 1
|
||||
}
|
||||
}
|
||||
|
||||
circuit-breakers.json: |
|
||||
{
|
||||
"dashboard": {
|
||||
"title": "Bakery IA - Circuit Breakers",
|
||||
"tags": ["bakery-ia", "reliability"],
|
||||
"timezone": "browser",
|
||||
"panels": [
|
||||
{
|
||||
"id": 1,
|
||||
"title": "Circuit Breaker States",
|
||||
"type": "stat",
|
||||
"gridPos": {"x": 0, "y": 0, "w": 24, "h": 4},
|
||||
"targets": [{
|
||||
"expr": "circuit_breaker_state",
|
||||
"legendFormat": "{{service}} - {{state}}"
|
||||
}]
|
||||
},
|
||||
{
|
||||
"id": 2,
|
||||
"title": "Circuit Breaker Trips",
|
||||
"type": "graph",
|
||||
"gridPos": {"x": 0, "y": 4, "w": 12, "h": 8},
|
||||
"targets": [{
|
||||
"expr": "rate(circuit_breaker_opened_total[5m])",
|
||||
"legendFormat": "{{service}}"
|
||||
}]
|
||||
},
|
||||
{
|
||||
"id": 3,
|
||||
"title": "Rejected Requests",
|
||||
"type": "graph",
|
||||
"gridPos": {"x": 12, "y": 4, "w": 12, "h": 8},
|
||||
"targets": [{
|
||||
"expr": "rate(circuit_breaker_rejected_total[5m])",
|
||||
"legendFormat": "{{service}}"
|
||||
}]
|
||||
}
|
||||
],
|
||||
"refresh": "10s",
|
||||
"schemaVersion": 16,
|
||||
"version": 1
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,146 @@
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: ConfigMap
|
||||
metadata:
|
||||
name: grafana-datasources
|
||||
namespace: monitoring
|
||||
data:
|
||||
prometheus.yaml: |
|
||||
apiVersion: 1
|
||||
datasources:
|
||||
- name: Prometheus
|
||||
type: prometheus
|
||||
access: proxy
|
||||
url: http://prometheus:9090
|
||||
isDefault: true
|
||||
editable: false
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: ConfigMap
|
||||
metadata:
|
||||
name: grafana-dashboards-config
|
||||
namespace: monitoring
|
||||
data:
|
||||
dashboards.yaml: |
|
||||
apiVersion: 1
|
||||
providers:
|
||||
- name: 'default'
|
||||
orgId: 1
|
||||
folder: 'Bakery IA'
|
||||
type: file
|
||||
disableDeletion: false
|
||||
updateIntervalSeconds: 10
|
||||
allowUiUpdates: true
|
||||
options:
|
||||
path: /var/lib/grafana/dashboards
|
||||
|
||||
---
|
||||
apiVersion: apps/v1
|
||||
kind: Deployment
|
||||
metadata:
|
||||
name: grafana
|
||||
namespace: monitoring
|
||||
labels:
|
||||
app: grafana
|
||||
spec:
|
||||
replicas: 1
|
||||
selector:
|
||||
matchLabels:
|
||||
app: grafana
|
||||
template:
|
||||
metadata:
|
||||
labels:
|
||||
app: grafana
|
||||
spec:
|
||||
containers:
|
||||
- name: grafana
|
||||
image: grafana/grafana:10.2.2
|
||||
ports:
|
||||
- containerPort: 3000
|
||||
name: http
|
||||
env:
|
||||
- name: GF_SECURITY_ADMIN_USER
|
||||
value: admin
|
||||
- name: GF_SECURITY_ADMIN_PASSWORD
|
||||
value: admin
|
||||
- name: GF_SERVER_ROOT_URL
|
||||
value: "http://monitoring.bakery-ia.local/grafana"
|
||||
- name: GF_SERVER_SERVE_FROM_SUB_PATH
|
||||
value: "true"
|
||||
- name: GF_AUTH_ANONYMOUS_ENABLED
|
||||
value: "false"
|
||||
- name: GF_INSTALL_PLUGINS
|
||||
value: ""
|
||||
volumeMounts:
|
||||
- name: grafana-storage
|
||||
mountPath: /var/lib/grafana
|
||||
- name: grafana-datasources
|
||||
mountPath: /etc/grafana/provisioning/datasources
|
||||
- name: grafana-dashboards-config
|
||||
mountPath: /etc/grafana/provisioning/dashboards
|
||||
- name: grafana-dashboards
|
||||
mountPath: /var/lib/grafana/dashboards
|
||||
resources:
|
||||
requests:
|
||||
memory: "256Mi"
|
||||
cpu: "100m"
|
||||
limits:
|
||||
memory: "512Mi"
|
||||
cpu: "500m"
|
||||
livenessProbe:
|
||||
httpGet:
|
||||
path: /api/health
|
||||
port: 3000
|
||||
initialDelaySeconds: 30
|
||||
periodSeconds: 10
|
||||
readinessProbe:
|
||||
httpGet:
|
||||
path: /api/health
|
||||
port: 3000
|
||||
initialDelaySeconds: 5
|
||||
periodSeconds: 5
|
||||
volumes:
|
||||
- name: grafana-storage
|
||||
persistentVolumeClaim:
|
||||
claimName: grafana-storage
|
||||
- name: grafana-datasources
|
||||
configMap:
|
||||
name: grafana-datasources
|
||||
- name: grafana-dashboards-config
|
||||
configMap:
|
||||
name: grafana-dashboards-config
|
||||
- name: grafana-dashboards
|
||||
configMap:
|
||||
name: grafana-dashboards
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: PersistentVolumeClaim
|
||||
metadata:
|
||||
name: grafana-storage
|
||||
namespace: monitoring
|
||||
spec:
|
||||
accessModes:
|
||||
- ReadWriteOnce
|
||||
resources:
|
||||
requests:
|
||||
storage: 5Gi
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: Service
|
||||
metadata:
|
||||
name: grafana
|
||||
namespace: monitoring
|
||||
labels:
|
||||
app: grafana
|
||||
spec:
|
||||
type: ClusterIP
|
||||
ports:
|
||||
- port: 3000
|
||||
targetPort: 3000
|
||||
protocol: TCP
|
||||
name: http
|
||||
selector:
|
||||
app: grafana
|
||||
@@ -0,0 +1,35 @@
|
||||
---
|
||||
apiVersion: networking.k8s.io/v1
|
||||
kind: Ingress
|
||||
metadata:
|
||||
name: monitoring-ingress
|
||||
namespace: monitoring
|
||||
annotations:
|
||||
nginx.ingress.kubernetes.io/rewrite-target: /$2
|
||||
nginx.ingress.kubernetes.io/ssl-redirect: "false"
|
||||
spec:
|
||||
rules:
|
||||
- host: monitoring.bakery-ia.local
|
||||
http:
|
||||
paths:
|
||||
- path: /grafana(/|$)(.*)
|
||||
pathType: ImplementationSpecific
|
||||
backend:
|
||||
service:
|
||||
name: grafana
|
||||
port:
|
||||
number: 3000
|
||||
- path: /prometheus(/|$)(.*)
|
||||
pathType: ImplementationSpecific
|
||||
backend:
|
||||
service:
|
||||
name: prometheus
|
||||
port:
|
||||
number: 9090
|
||||
- path: /jaeger(/|$)(.*)
|
||||
pathType: ImplementationSpecific
|
||||
backend:
|
||||
service:
|
||||
name: jaeger-query
|
||||
port:
|
||||
number: 16686
|
||||
190
infrastructure/kubernetes/base/components/monitoring/jaeger.yaml
Normal file
190
infrastructure/kubernetes/base/components/monitoring/jaeger.yaml
Normal file
@@ -0,0 +1,190 @@
|
||||
---
|
||||
apiVersion: apps/v1
|
||||
kind: Deployment
|
||||
metadata:
|
||||
name: jaeger
|
||||
namespace: monitoring
|
||||
labels:
|
||||
app: jaeger
|
||||
spec:
|
||||
replicas: 1
|
||||
selector:
|
||||
matchLabels:
|
||||
app: jaeger
|
||||
template:
|
||||
metadata:
|
||||
labels:
|
||||
app: jaeger
|
||||
spec:
|
||||
containers:
|
||||
- name: jaeger
|
||||
image: jaegertracing/all-in-one:1.51
|
||||
env:
|
||||
- name: COLLECTOR_ZIPKIN_HOST_PORT
|
||||
value: ":9411"
|
||||
- name: COLLECTOR_OTLP_ENABLED
|
||||
value: "true"
|
||||
- name: SPAN_STORAGE_TYPE
|
||||
value: "badger"
|
||||
- name: BADGER_EPHEMERAL
|
||||
value: "false"
|
||||
- name: BADGER_DIRECTORY_VALUE
|
||||
value: "/badger/data"
|
||||
- name: BADGER_DIRECTORY_KEY
|
||||
value: "/badger/key"
|
||||
ports:
|
||||
- containerPort: 5775
|
||||
protocol: UDP
|
||||
name: zipkin-compact
|
||||
- containerPort: 6831
|
||||
protocol: UDP
|
||||
name: jaeger-compact
|
||||
- containerPort: 6832
|
||||
protocol: UDP
|
||||
name: jaeger-binary
|
||||
- containerPort: 5778
|
||||
protocol: TCP
|
||||
name: config-rest
|
||||
- containerPort: 16686
|
||||
protocol: TCP
|
||||
name: query
|
||||
- containerPort: 14250
|
||||
protocol: TCP
|
||||
name: grpc
|
||||
- containerPort: 14268
|
||||
protocol: TCP
|
||||
name: c-tchan-trft
|
||||
- containerPort: 14269
|
||||
protocol: TCP
|
||||
name: admin-http
|
||||
- containerPort: 9411
|
||||
protocol: TCP
|
||||
name: zipkin
|
||||
- containerPort: 4317
|
||||
protocol: TCP
|
||||
name: otlp-grpc
|
||||
- containerPort: 4318
|
||||
protocol: TCP
|
||||
name: otlp-http
|
||||
volumeMounts:
|
||||
- name: jaeger-storage
|
||||
mountPath: /badger
|
||||
resources:
|
||||
requests:
|
||||
memory: "512Mi"
|
||||
cpu: "250m"
|
||||
limits:
|
||||
memory: "1Gi"
|
||||
cpu: "500m"
|
||||
livenessProbe:
|
||||
httpGet:
|
||||
path: /
|
||||
port: 14269
|
||||
initialDelaySeconds: 30
|
||||
periodSeconds: 10
|
||||
readinessProbe:
|
||||
httpGet:
|
||||
path: /
|
||||
port: 14269
|
||||
initialDelaySeconds: 5
|
||||
periodSeconds: 5
|
||||
volumes:
|
||||
- name: jaeger-storage
|
||||
persistentVolumeClaim:
|
||||
claimName: jaeger-storage
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: PersistentVolumeClaim
|
||||
metadata:
|
||||
name: jaeger-storage
|
||||
namespace: monitoring
|
||||
spec:
|
||||
accessModes:
|
||||
- ReadWriteOnce
|
||||
resources:
|
||||
requests:
|
||||
storage: 10Gi
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: Service
|
||||
metadata:
|
||||
name: jaeger-query
|
||||
namespace: monitoring
|
||||
labels:
|
||||
app: jaeger
|
||||
spec:
|
||||
type: ClusterIP
|
||||
ports:
|
||||
- port: 16686
|
||||
targetPort: 16686
|
||||
protocol: TCP
|
||||
name: query
|
||||
selector:
|
||||
app: jaeger
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: Service
|
||||
metadata:
|
||||
name: jaeger-collector
|
||||
namespace: monitoring
|
||||
labels:
|
||||
app: jaeger
|
||||
spec:
|
||||
type: ClusterIP
|
||||
ports:
|
||||
- port: 14268
|
||||
targetPort: 14268
|
||||
protocol: TCP
|
||||
name: c-tchan-trft
|
||||
- port: 14250
|
||||
targetPort: 14250
|
||||
protocol: TCP
|
||||
name: grpc
|
||||
- port: 9411
|
||||
targetPort: 9411
|
||||
protocol: TCP
|
||||
name: zipkin
|
||||
- port: 4317
|
||||
targetPort: 4317
|
||||
protocol: TCP
|
||||
name: otlp-grpc
|
||||
- port: 4318
|
||||
targetPort: 4318
|
||||
protocol: TCP
|
||||
name: otlp-http
|
||||
selector:
|
||||
app: jaeger
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: Service
|
||||
metadata:
|
||||
name: jaeger-agent
|
||||
namespace: monitoring
|
||||
labels:
|
||||
app: jaeger
|
||||
spec:
|
||||
type: ClusterIP
|
||||
clusterIP: None
|
||||
ports:
|
||||
- port: 5775
|
||||
targetPort: 5775
|
||||
protocol: UDP
|
||||
name: zipkin-compact
|
||||
- port: 6831
|
||||
targetPort: 6831
|
||||
protocol: UDP
|
||||
name: jaeger-compact
|
||||
- port: 6832
|
||||
targetPort: 6832
|
||||
protocol: UDP
|
||||
name: jaeger-binary
|
||||
- port: 5778
|
||||
targetPort: 5778
|
||||
protocol: TCP
|
||||
name: config-rest
|
||||
selector:
|
||||
app: jaeger
|
||||
@@ -0,0 +1,10 @@
|
||||
apiVersion: kustomize.config.k8s.io/v1beta1
|
||||
kind: Kustomization
|
||||
|
||||
resources:
|
||||
- namespace.yaml
|
||||
- prometheus.yaml
|
||||
- grafana.yaml
|
||||
- grafana-dashboards.yaml
|
||||
- jaeger.yaml
|
||||
- ingress.yaml
|
||||
@@ -0,0 +1,7 @@
|
||||
apiVersion: v1
|
||||
kind: Namespace
|
||||
metadata:
|
||||
name: monitoring
|
||||
labels:
|
||||
name: monitoring
|
||||
app.kubernetes.io/part-of: bakery-ia
|
||||
@@ -0,0 +1,210 @@
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: ServiceAccount
|
||||
metadata:
|
||||
name: prometheus
|
||||
namespace: monitoring
|
||||
|
||||
---
|
||||
apiVersion: rbac.authorization.k8s.io/v1
|
||||
kind: ClusterRole
|
||||
metadata:
|
||||
name: prometheus
|
||||
rules:
|
||||
- apiGroups: [""]
|
||||
resources:
|
||||
- nodes
|
||||
- nodes/proxy
|
||||
- services
|
||||
- endpoints
|
||||
- pods
|
||||
verbs: ["get", "list", "watch"]
|
||||
- apiGroups:
|
||||
- extensions
|
||||
resources:
|
||||
- ingresses
|
||||
verbs: ["get", "list", "watch"]
|
||||
- nonResourceURLs: ["/metrics"]
|
||||
verbs: ["get"]
|
||||
|
||||
---
|
||||
apiVersion: rbac.authorization.k8s.io/v1
|
||||
kind: ClusterRoleBinding
|
||||
metadata:
|
||||
name: prometheus
|
||||
roleRef:
|
||||
apiGroup: rbac.authorization.k8s.io
|
||||
kind: ClusterRole
|
||||
name: prometheus
|
||||
subjects:
|
||||
- kind: ServiceAccount
|
||||
name: prometheus
|
||||
namespace: monitoring
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: ConfigMap
|
||||
metadata:
|
||||
name: prometheus-config
|
||||
namespace: monitoring
|
||||
data:
|
||||
prometheus.yml: |
|
||||
global:
|
||||
scrape_interval: 30s
|
||||
evaluation_interval: 30s
|
||||
external_labels:
|
||||
cluster: 'bakery-ia'
|
||||
environment: 'production'
|
||||
|
||||
scrape_configs:
|
||||
# Scrape Prometheus itself
|
||||
- job_name: 'prometheus'
|
||||
static_configs:
|
||||
- targets: ['localhost:9090']
|
||||
|
||||
# Scrape all bakery-ia services
|
||||
- job_name: 'bakery-services'
|
||||
kubernetes_sd_configs:
|
||||
- role: pod
|
||||
namespaces:
|
||||
names:
|
||||
- bakery-ia
|
||||
relabel_configs:
|
||||
# Only scrape pods with metrics port
|
||||
- source_labels: [__meta_kubernetes_pod_container_port_name]
|
||||
action: keep
|
||||
regex: http
|
||||
|
||||
# Add service name label
|
||||
- source_labels: [__meta_kubernetes_pod_label_app_kubernetes_io_name]
|
||||
target_label: service
|
||||
|
||||
# Add component label
|
||||
- source_labels: [__meta_kubernetes_pod_label_app_kubernetes_io_component]
|
||||
target_label: component
|
||||
|
||||
# Add pod name
|
||||
- source_labels: [__meta_kubernetes_pod_name]
|
||||
target_label: pod
|
||||
|
||||
# Set metrics path
|
||||
- source_labels: [__meta_kubernetes_pod_annotation_prometheus_io_path]
|
||||
action: replace
|
||||
target_label: __metrics_path__
|
||||
regex: (.+)
|
||||
|
||||
# Set scrape port
|
||||
- source_labels: [__address__, __meta_kubernetes_pod_annotation_prometheus_io_port]
|
||||
action: replace
|
||||
regex: ([^:]+)(?::\d+)?;(\d+)
|
||||
replacement: $1:$2
|
||||
target_label: __address__
|
||||
|
||||
# Scrape Kubernetes nodes
|
||||
- job_name: 'kubernetes-nodes'
|
||||
kubernetes_sd_configs:
|
||||
- role: node
|
||||
relabel_configs:
|
||||
- action: labelmap
|
||||
regex: __meta_kubernetes_node_label_(.+)
|
||||
- target_label: __address__
|
||||
replacement: kubernetes.default.svc:443
|
||||
- source_labels: [__meta_kubernetes_node_name]
|
||||
regex: (.+)
|
||||
target_label: __metrics_path__
|
||||
replacement: /api/v1/nodes/${1}/proxy/metrics
|
||||
|
||||
---
|
||||
apiVersion: apps/v1
|
||||
kind: Deployment
|
||||
metadata:
|
||||
name: prometheus
|
||||
namespace: monitoring
|
||||
labels:
|
||||
app: prometheus
|
||||
spec:
|
||||
replicas: 1
|
||||
selector:
|
||||
matchLabels:
|
||||
app: prometheus
|
||||
template:
|
||||
metadata:
|
||||
labels:
|
||||
app: prometheus
|
||||
spec:
|
||||
serviceAccountName: prometheus
|
||||
containers:
|
||||
- name: prometheus
|
||||
image: prom/prometheus:v2.48.0
|
||||
args:
|
||||
- '--config.file=/etc/prometheus/prometheus.yml'
|
||||
- '--storage.tsdb.path=/prometheus'
|
||||
- '--storage.tsdb.retention.time=30d'
|
||||
- '--web.console.libraries=/usr/share/prometheus/console_libraries'
|
||||
- '--web.console.templates=/usr/share/prometheus/consoles'
|
||||
- '--web.enable-lifecycle'
|
||||
ports:
|
||||
- containerPort: 9090
|
||||
name: web
|
||||
volumeMounts:
|
||||
- name: prometheus-config
|
||||
mountPath: /etc/prometheus
|
||||
- name: prometheus-storage
|
||||
mountPath: /prometheus
|
||||
resources:
|
||||
requests:
|
||||
memory: "1Gi"
|
||||
cpu: "500m"
|
||||
limits:
|
||||
memory: "2Gi"
|
||||
cpu: "1"
|
||||
livenessProbe:
|
||||
httpGet:
|
||||
path: /-/healthy
|
||||
port: 9090
|
||||
initialDelaySeconds: 30
|
||||
periodSeconds: 10
|
||||
readinessProbe:
|
||||
httpGet:
|
||||
path: /-/ready
|
||||
port: 9090
|
||||
initialDelaySeconds: 5
|
||||
periodSeconds: 5
|
||||
volumes:
|
||||
- name: prometheus-config
|
||||
configMap:
|
||||
name: prometheus-config
|
||||
- name: prometheus-storage
|
||||
persistentVolumeClaim:
|
||||
claimName: prometheus-storage
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: PersistentVolumeClaim
|
||||
metadata:
|
||||
name: prometheus-storage
|
||||
namespace: monitoring
|
||||
spec:
|
||||
accessModes:
|
||||
- ReadWriteOnce
|
||||
resources:
|
||||
requests:
|
||||
storage: 20Gi
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: Service
|
||||
metadata:
|
||||
name: prometheus
|
||||
namespace: monitoring
|
||||
labels:
|
||||
app: prometheus
|
||||
spec:
|
||||
type: ClusterIP
|
||||
ports:
|
||||
- port: 9090
|
||||
targetPort: 9090
|
||||
protocol: TCP
|
||||
name: web
|
||||
selector:
|
||||
app: prometheus
|
||||
@@ -0,0 +1,158 @@
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: ConfigMap
|
||||
metadata:
|
||||
name: nominatim-config
|
||||
namespace: bakery-ia
|
||||
labels:
|
||||
app.kubernetes.io/name: nominatim
|
||||
app.kubernetes.io/component: geocoding
|
||||
data:
|
||||
NOMINATIM_PBF_URL: "http://download.geofabrik.de/europe/spain-latest.osm.pbf"
|
||||
NOMINATIM_REPLICATION_URL: "https://download.geofabrik.de/europe/spain-updates"
|
||||
NOMINATIM_IMPORT_STYLE: "address"
|
||||
NOMINATIM_THREADS: "4"
|
||||
NOMINATIM_FLATNODE_FILE: "/nominatim-flatnode/flatnode.bin"
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: PersistentVolumeClaim
|
||||
metadata:
|
||||
name: nominatim-data
|
||||
namespace: bakery-ia
|
||||
labels:
|
||||
app.kubernetes.io/name: nominatim
|
||||
app.kubernetes.io/component: geocoding
|
||||
spec:
|
||||
accessModes:
|
||||
- ReadWriteOnce
|
||||
resources:
|
||||
requests:
|
||||
storage: 50Gi
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: PersistentVolumeClaim
|
||||
metadata:
|
||||
name: nominatim-flatnode
|
||||
namespace: bakery-ia
|
||||
labels:
|
||||
app.kubernetes.io/name: nominatim
|
||||
app.kubernetes.io/component: geocoding
|
||||
spec:
|
||||
accessModes:
|
||||
- ReadWriteOnce
|
||||
resources:
|
||||
requests:
|
||||
storage: 20Gi
|
||||
|
||||
---
|
||||
apiVersion: apps/v1
|
||||
kind: StatefulSet
|
||||
metadata:
|
||||
name: nominatim
|
||||
namespace: bakery-ia
|
||||
labels:
|
||||
app.kubernetes.io/name: nominatim
|
||||
app.kubernetes.io/component: geocoding
|
||||
app.kubernetes.io/part-of: bakery-ia
|
||||
spec:
|
||||
serviceName: nominatim-service
|
||||
replicas: 1
|
||||
selector:
|
||||
matchLabels:
|
||||
app.kubernetes.io/name: nominatim
|
||||
app.kubernetes.io/component: geocoding
|
||||
template:
|
||||
metadata:
|
||||
labels:
|
||||
app.kubernetes.io/name: nominatim
|
||||
app.kubernetes.io/component: geocoding
|
||||
spec:
|
||||
containers:
|
||||
- name: nominatim
|
||||
image: mediagis/nominatim:4.4
|
||||
ports:
|
||||
- containerPort: 8080
|
||||
name: http
|
||||
volumeMounts:
|
||||
- name: nominatim-data
|
||||
mountPath: /var/lib/postgresql
|
||||
- name: nominatim-flatnode
|
||||
mountPath: /nominatim-flatnode
|
||||
env:
|
||||
- name: NOMINATIM_PBF_URL
|
||||
valueFrom:
|
||||
configMapKeyRef:
|
||||
name: nominatim-config
|
||||
key: NOMINATIM_PBF_URL
|
||||
- name: NOMINATIM_REPLICATION_URL
|
||||
valueFrom:
|
||||
configMapKeyRef:
|
||||
name: nominatim-config
|
||||
key: NOMINATIM_REPLICATION_URL
|
||||
- name: NOMINATIM_IMPORT_STYLE
|
||||
valueFrom:
|
||||
configMapKeyRef:
|
||||
name: nominatim-config
|
||||
key: NOMINATIM_IMPORT_STYLE
|
||||
- name: NOMINATIM_THREADS
|
||||
valueFrom:
|
||||
configMapKeyRef:
|
||||
name: nominatim-config
|
||||
key: NOMINATIM_THREADS
|
||||
- name: NOMINATIM_FLATNODE_FILE
|
||||
valueFrom:
|
||||
configMapKeyRef:
|
||||
name: nominatim-config
|
||||
key: NOMINATIM_FLATNODE_FILE
|
||||
resources:
|
||||
requests:
|
||||
memory: "2Gi"
|
||||
cpu: "1"
|
||||
limits:
|
||||
memory: "4Gi"
|
||||
cpu: "2"
|
||||
livenessProbe:
|
||||
httpGet:
|
||||
path: /status
|
||||
port: 8080
|
||||
initialDelaySeconds: 120
|
||||
periodSeconds: 30
|
||||
timeoutSeconds: 10
|
||||
failureThreshold: 3
|
||||
readinessProbe:
|
||||
httpGet:
|
||||
path: /status
|
||||
port: 8080
|
||||
initialDelaySeconds: 60
|
||||
periodSeconds: 10
|
||||
timeoutSeconds: 5
|
||||
failureThreshold: 5
|
||||
volumes:
|
||||
- name: nominatim-data
|
||||
persistentVolumeClaim:
|
||||
claimName: nominatim-data
|
||||
- name: nominatim-flatnode
|
||||
persistentVolumeClaim:
|
||||
claimName: nominatim-flatnode
|
||||
|
||||
---
|
||||
apiVersion: v1
|
||||
kind: Service
|
||||
metadata:
|
||||
name: nominatim-service
|
||||
namespace: bakery-ia
|
||||
labels:
|
||||
app.kubernetes.io/name: nominatim
|
||||
app.kubernetes.io/component: geocoding
|
||||
spec:
|
||||
selector:
|
||||
app.kubernetes.io/name: nominatim
|
||||
app.kubernetes.io/component: geocoding
|
||||
ports:
|
||||
- port: 8080
|
||||
targetPort: 8080
|
||||
protocol: TCP
|
||||
name: http
|
||||
type: ClusterIP
|
||||
@@ -2,6 +2,7 @@ apiVersion: v1
|
||||
kind: ConfigMap
|
||||
metadata:
|
||||
name: bakery-config
|
||||
namespace: bakery-ia
|
||||
labels:
|
||||
app.kubernetes.io/name: bakery-ia
|
||||
app.kubernetes.io/component: config
|
||||
@@ -9,7 +10,7 @@ data:
|
||||
# ================================================================
|
||||
# ENVIRONMENT & BUILD SETTINGS
|
||||
# ================================================================
|
||||
ENVIRONMENT: "production"
|
||||
ENVIRONMENT: "development"
|
||||
DEBUG: "false"
|
||||
LOG_LEVEL: "INFO"
|
||||
|
||||
@@ -323,12 +324,22 @@ data:
|
||||
VITE_ENVIRONMENT: "production"
|
||||
|
||||
# ================================================================
|
||||
# LOCATION SETTINGS
|
||||
# LOCATION SETTINGS (Nominatim Geocoding)
|
||||
# ================================================================
|
||||
NOMINATIM_SERVICE_URL: "http://nominatim-service:8080"
|
||||
NOMINATIM_PBF_URL: "http://download.geofabrik.de/europe/spain-latest.osm.pbf"
|
||||
NOMINATIM_MEMORY_LIMIT: "8G"
|
||||
NOMINATIM_CPU_LIMIT: "4"
|
||||
|
||||
# ================================================================
|
||||
# DISTRIBUTED TRACING (Jaeger/OpenTelemetry)
|
||||
# ================================================================
|
||||
JAEGER_COLLECTOR_ENDPOINT: "http://jaeger-collector.monitoring:4317"
|
||||
JAEGER_AGENT_HOST: "jaeger-agent.monitoring"
|
||||
JAEGER_AGENT_PORT: "6831"
|
||||
OTEL_EXPORTER_OTLP_ENDPOINT: "http://jaeger-collector.monitoring:4317"
|
||||
OTEL_SERVICE_NAME: "bakery-ia"
|
||||
|
||||
# ================================================================
|
||||
# EXTERNAL DATA SERVICE V2 SETTINGS
|
||||
# ================================================================
|
||||
|
||||
@@ -2,6 +2,7 @@ apiVersion: v1
|
||||
kind: ConfigMap
|
||||
metadata:
|
||||
name: postgres-init-config
|
||||
namespace: bakery-ia
|
||||
labels:
|
||||
app.kubernetes.io/component: database
|
||||
app.kubernetes.io/part-of: bakery-ia
|
||||
|
||||
83
infrastructure/kubernetes/base/jobs/nominatim-init-job.yaml
Normal file
83
infrastructure/kubernetes/base/jobs/nominatim-init-job.yaml
Normal file
@@ -0,0 +1,83 @@
|
||||
apiVersion: batch/v1
|
||||
kind: Job
|
||||
metadata:
|
||||
name: nominatim-init
|
||||
namespace: bakery-ia
|
||||
labels:
|
||||
app.kubernetes.io/name: nominatim-init
|
||||
app.kubernetes.io/component: data-init
|
||||
app.kubernetes.io/part-of: bakery-ia
|
||||
spec:
|
||||
ttlSecondsAfterFinished: 86400
|
||||
template:
|
||||
metadata:
|
||||
labels:
|
||||
app.kubernetes.io/name: nominatim-init
|
||||
app.kubernetes.io/component: data-init
|
||||
spec:
|
||||
restartPolicy: OnFailure
|
||||
containers:
|
||||
- name: nominatim-import
|
||||
image: mediagis/nominatim:4.4
|
||||
command:
|
||||
- sh
|
||||
- -c
|
||||
- |
|
||||
set -e
|
||||
echo "Checking if Nominatim database is already initialized..."
|
||||
|
||||
if psql -lqt | cut -d \| -f 1 | grep -qw nominatim; then
|
||||
echo "Nominatim database already exists. Skipping import."
|
||||
exit 0
|
||||
fi
|
||||
|
||||
echo "Downloading Spain OSM data..."
|
||||
wget -O /tmp/spain-latest.osm.pbf "${NOMINATIM_PBF_URL}"
|
||||
|
||||
echo "Importing OSM data into Nominatim (this may take 30-60 minutes)..."
|
||||
nominatim import --osm-file /tmp/spain-latest.osm.pbf
|
||||
|
||||
echo "Building search indices..."
|
||||
nominatim refresh --website --importance
|
||||
|
||||
echo "Nominatim initialization complete!"
|
||||
volumeMounts:
|
||||
- name: nominatim-data
|
||||
mountPath: /var/lib/postgresql
|
||||
- name: nominatim-flatnode
|
||||
mountPath: /nominatim-flatnode
|
||||
env:
|
||||
- name: NOMINATIM_PBF_URL
|
||||
valueFrom:
|
||||
configMapKeyRef:
|
||||
name: nominatim-config
|
||||
key: NOMINATIM_PBF_URL
|
||||
- name: NOMINATIM_IMPORT_STYLE
|
||||
valueFrom:
|
||||
configMapKeyRef:
|
||||
name: nominatim-config
|
||||
key: NOMINATIM_IMPORT_STYLE
|
||||
- name: NOMINATIM_THREADS
|
||||
valueFrom:
|
||||
configMapKeyRef:
|
||||
name: nominatim-config
|
||||
key: NOMINATIM_THREADS
|
||||
- name: NOMINATIM_FLATNODE_FILE
|
||||
valueFrom:
|
||||
configMapKeyRef:
|
||||
name: nominatim-config
|
||||
key: NOMINATIM_FLATNODE_FILE
|
||||
resources:
|
||||
requests:
|
||||
memory: "8Gi"
|
||||
cpu: "4"
|
||||
limits:
|
||||
memory: "16Gi"
|
||||
cpu: "8"
|
||||
volumes:
|
||||
- name: nominatim-data
|
||||
persistentVolumeClaim:
|
||||
claimName: nominatim-data
|
||||
- name: nominatim-flatnode
|
||||
persistentVolumeClaim:
|
||||
claimName: nominatim-flatnode
|
||||
@@ -55,6 +55,10 @@ resources:
|
||||
- components/databases/rabbitmq.yaml
|
||||
- components/infrastructure/gateway-service.yaml
|
||||
|
||||
# Nominatim geocoding service
|
||||
- components/nominatim/nominatim.yaml
|
||||
- jobs/nominatim-init-job.yaml
|
||||
|
||||
# Persistent storage
|
||||
- components/volumes/model-storage-pvc.yaml
|
||||
|
||||
|
||||
@@ -4,13 +4,101 @@ kind: Kustomization
|
||||
metadata:
|
||||
name: bakery-ia-dev
|
||||
|
||||
namespace: bakery-ia
|
||||
# Note: Removed global namespace to prevent monitoring namespace conflict
|
||||
# All base resources already have namespace: bakery-ia defined
|
||||
|
||||
resources:
|
||||
- ../../base
|
||||
# Monitoring disabled for dev to save resources
|
||||
# - ../../base/components/monitoring
|
||||
- dev-ingress.yaml
|
||||
|
||||
# Exclude nominatim from dev to save resources
|
||||
# Using scale to 0 for StatefulSet to prevent pod creation
|
||||
patches:
|
||||
# Override specific ConfigMap values for development
|
||||
- target:
|
||||
kind: ConfigMap
|
||||
name: bakery-config
|
||||
patch: |-
|
||||
- op: replace
|
||||
path: /data/ENVIRONMENT
|
||||
value: "development"
|
||||
- op: replace
|
||||
path: /data/DEBUG
|
||||
value: "true"
|
||||
- op: replace
|
||||
path: /data/LOG_LEVEL
|
||||
value: "DEBUG"
|
||||
- op: replace
|
||||
path: /data/AUTO_RELOAD
|
||||
value: "true"
|
||||
- op: replace
|
||||
path: /data/PROFILING_ENABLED
|
||||
value: "true"
|
||||
- op: replace
|
||||
path: /data/MOCK_EXTERNAL_APIS
|
||||
value: "true"
|
||||
- op: replace
|
||||
path: /data/TESTING
|
||||
value: "false"
|
||||
- op: replace
|
||||
path: /data/DOMAIN
|
||||
value: "localhost"
|
||||
- op: replace
|
||||
path: /data/API_DOCS_ENABLED
|
||||
value: "true"
|
||||
- op: replace
|
||||
path: /data/CORS_ORIGINS
|
||||
value: "http://frontend-service:3000,http://localhost:3000,http://localhost:3001,http://localhost,http://127.0.0.1:3000,http://127.0.0.1:3001,http://bakery-ia.local,https://localhost,https://127.0.0.1"
|
||||
- op: replace
|
||||
path: /data/VITE_ENVIRONMENT
|
||||
value: "development"
|
||||
- op: replace
|
||||
path: /data/VITE_API_URL
|
||||
value: "/api"
|
||||
- op: replace
|
||||
path: /data/STRIPE_PUBLISHABLE_KEY
|
||||
value: "pk_test_your_stripe_publishable_key_here"
|
||||
- op: replace
|
||||
path: /data/SQUARE_ENVIRONMENT
|
||||
value: "sandbox"
|
||||
- op: replace
|
||||
path: /data/TOAST_ENVIRONMENT
|
||||
value: "sandbox"
|
||||
- op: replace
|
||||
path: /data/LIGHTSPEED_ENVIRONMENT
|
||||
value: "sandbox"
|
||||
- op: replace
|
||||
path: /data/RATE_LIMIT_ENABLED
|
||||
value: "false"
|
||||
- op: replace
|
||||
path: /data/DB_FORCE_RECREATE
|
||||
value: "false"
|
||||
- op: add
|
||||
path: /data/DEVELOPMENT_MODE
|
||||
value: "true"
|
||||
- op: add
|
||||
path: /data/DEBUG_LOGGING
|
||||
value: "true"
|
||||
- op: add
|
||||
path: /data/SKIP_MIGRATION_VERSION_CHECK
|
||||
value: "false"
|
||||
- target:
|
||||
kind: StatefulSet
|
||||
name: nominatim
|
||||
patch: |-
|
||||
- op: replace
|
||||
path: /spec/replicas
|
||||
value: 0
|
||||
# Suspend nominatim-init job in dev (not needed when nominatim is scaled to 0)
|
||||
- target:
|
||||
kind: Job
|
||||
name: nominatim-init
|
||||
patch: |-
|
||||
- op: replace
|
||||
path: /spec/suspend
|
||||
value: true
|
||||
- target:
|
||||
group: apps
|
||||
version: v1
|
||||
@@ -485,43 +573,6 @@ patches:
|
||||
memory: "1Gi"
|
||||
cpu: "500m"
|
||||
|
||||
configMapGenerator:
|
||||
- name: bakery-config
|
||||
behavior: merge
|
||||
literals:
|
||||
# Environment & Build Settings
|
||||
- ENVIRONMENT=development
|
||||
- DEBUG=true
|
||||
- LOG_LEVEL=DEBUG
|
||||
- AUTO_RELOAD=true
|
||||
- PROFILING_ENABLED=true
|
||||
- MOCK_EXTERNAL_APIS=true
|
||||
- TESTING=false
|
||||
- DOMAIN=localhost
|
||||
- API_DOCS_ENABLED=true
|
||||
|
||||
# CORS Configuration for Development
|
||||
- CORS_ORIGINS=http://frontend-service:3000,http://localhost:3000,http://localhost:3001,http://localhost,http://127.0.0.1:3000,http://127.0.0.1:3001,http://bakery-ia.local,https://localhost,https://127.0.0.1
|
||||
|
||||
# Frontend Development Configuration
|
||||
- VITE_ENVIRONMENT=development
|
||||
- VITE_API_URL=/api
|
||||
|
||||
# Payment Configuration (Sandbox for dev)
|
||||
- STRIPE_PUBLISHABLE_KEY=pk_test_your_stripe_publishable_key_here
|
||||
- SQUARE_ENVIRONMENT=sandbox
|
||||
- TOAST_ENVIRONMENT=sandbox
|
||||
- LIGHTSPEED_ENVIRONMENT=sandbox
|
||||
|
||||
# Rate Limiting (Disabled for dev)
|
||||
- RATE_LIMIT_ENABLED=false
|
||||
|
||||
# Database (Development mode)
|
||||
- DB_FORCE_RECREATE=false
|
||||
- DEVELOPMENT_MODE=true
|
||||
- DEBUG_LOGGING=true
|
||||
- SKIP_MIGRATION_VERSION_CHECK=false
|
||||
|
||||
secretGenerator:
|
||||
- name: dev-secrets
|
||||
literals:
|
||||
|
||||
Reference in New Issue
Block a user