declarative relay deployment on hetzner
relay-eval.waow.tech
atproto
relay
fix: separate monitoring values for relay and zlay clusters
each cluster was getting both dashboards (relay + zlay) even though
it can only see its own metrics. split monitoring-values.yaml into
relay-specific and zlay-specific files. also add zlay-dashboard
configmap creation to zlay-deploy recipe.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
+91
-11
-9
deploy/monitoring-values.yaml
-9
deploy/monitoring-values.yaml
···
72
72
editable: true
73
73
options:
74
74
path: /var/lib/grafana/dashboards/relay
75
-
- name: zlay
76
-
orgId: 1
77
-
folder: ""
78
-
type: file
79
-
disableDeletion: false
80
-
editable: true
81
-
options:
82
-
path: /var/lib/grafana/dashboards/zlay
83
75
dashboardsConfigMaps:
84
76
relay: relay-dashboard
85
-
zlay: zlay-dashboard
86
77
defaultDashboardsEnabled: false
87
78
88
79
# --- operator ---
+86
deploy/zlay-monitoring-values.yaml
+86
deploy/zlay-monitoring-values.yaml
···
1
+
# kube-prometheus-stack helm values for the zlay cluster
2
+
# docs: https://github.com/prometheus-community/helm-charts/tree/main/charts/kube-prometheus-stack
3
+
4
+
# --- disable everything we don't need ---
5
+
alertmanager:
6
+
enabled: false
7
+
nodeExporter:
8
+
enabled: false
9
+
kubeStateMetrics:
10
+
enabled: false
11
+
kubeApiServer:
12
+
enabled: false
13
+
kubeControllerManager:
14
+
enabled: false
15
+
kubeScheduler:
16
+
enabled: false
17
+
kubeProxy:
18
+
enabled: false
19
+
kubeEtcd:
20
+
enabled: false
21
+
kubeDns:
22
+
enabled: false
23
+
coreDns:
24
+
enabled: false
25
+
kubelet:
26
+
enabled: true
27
+
28
+
# --- prometheus ---
29
+
prometheus:
30
+
prometheusSpec:
31
+
scrapeInterval: 30s
32
+
retention: 14d
33
+
resources:
34
+
requests:
35
+
memory: 256Mi
36
+
cpu: 100m
37
+
limits:
38
+
memory: 1Gi
39
+
serviceMonitorSelectorNilUsesHelmValues: false
40
+
storageSpec:
41
+
volumeClaimTemplate:
42
+
spec:
43
+
accessModes: ["ReadWriteOnce"]
44
+
resources:
45
+
requests:
46
+
storage: 10Gi
47
+
48
+
# --- grafana ---
49
+
grafana:
50
+
resources:
51
+
requests:
52
+
memory: 128Mi
53
+
cpu: 50m
54
+
limits:
55
+
memory: 256Mi
56
+
grafana.ini:
57
+
auth.anonymous:
58
+
enabled: true
59
+
org_role: Viewer
60
+
server:
61
+
root_url: https://zlay-metrics.waow.tech
62
+
adminPassword: ${GRAFANA_ADMIN_PASSWORD:-prom-operator}
63
+
dashboardProviders:
64
+
dashboardproviders.yaml:
65
+
apiVersion: 1
66
+
providers:
67
+
- name: zlay
68
+
orgId: 1
69
+
folder: ""
70
+
type: file
71
+
disableDeletion: false
72
+
editable: true
73
+
options:
74
+
path: /var/lib/grafana/dashboards/zlay
75
+
dashboardsConfigMaps:
76
+
zlay: zlay-dashboard
77
+
defaultDashboardsEnabled: false
78
+
79
+
# --- operator ---
80
+
prometheusOperator:
81
+
resources:
82
+
requests:
83
+
memory: 64Mi
84
+
cpu: 10m
85
+
limits:
86
+
memory: 128Mi
+5
-2
justfile
+5
-2
justfile
···
437
437
echo "==> installing monitoring"
438
438
ZLAY_METRICS_DOMAIN="${ZLAY_METRICS_DOMAIN:-zlay-metrics.waow.tech}"
439
439
kubectl create namespace monitoring --dry-run=client -o yaml | kubectl apply -f -
440
+
kubectl create configmap zlay-dashboard \
441
+
--namespace monitoring \
442
+
--from-file=zlay-dashboard.json=deploy/zlay-dashboard.json \
443
+
--dry-run=client -o yaml | kubectl apply -f -
440
444
helm upgrade --install kube-prometheus-stack prometheus-community/kube-prometheus-stack \
441
445
--namespace monitoring \
442
-
--values deploy/monitoring-values.yaml \
446
+
--values deploy/zlay-monitoring-values.yaml \
443
447
--set grafana.adminPassword="${GRAFANA_ADMIN_PASSWORD:-prom-operator}" \
444
-
--set "grafana.grafana\\.ini.server.root_url=https://$ZLAY_METRICS_DOMAIN" \
445
448
--wait --timeout 5m
446
449
kubectl apply -f deploy/zlay-servicemonitor.yaml
447
450