Skip to content

Commit 9ef10f6

Browse files
committed
Add Aria scrape config, Grafana LB, disable node-exporter
- Add additionalScrapeConfigs for aria-streaming (Spark :5182) and k3s-node-exporter (all 4 nodes on :9100) - Expose Grafana via LoadBalancer at 10.10.88.86 - Disable bundled node-exporter (npu-device-monitor already provides node_exporter on hostPort 9100) - Remove dashboardProviders config (sidecar handles it)
1 parent e912763 commit 9ef10f6

File tree

1 file changed

+20
-14
lines changed

1 file changed

+20
-14
lines changed

ansible/templates/prometheus-values.yml.j2

Lines changed: 20 additions & 14 deletions
Original file line numberDiff line numberDiff line change
@@ -23,6 +23,20 @@ prometheus:
2323
# ARM64 optimizations
2424
podMonitorSelectorNilUsesHelmValues: false
2525
serviceMonitorSelectorNilUsesHelmValues: false
26+
additionalScrapeConfigs:
27+
- job_name: 'aria-streaming'
28+
static_configs:
29+
- targets: ['10.10.88.120:5182']
30+
metrics_path: '/api/metrics/prometheus'
31+
scrape_interval: 30s
32+
- job_name: 'k3s-node-exporter'
33+
static_configs:
34+
- targets:
35+
- '10.10.88.73:9100'
36+
- '10.10.88.74:9100'
37+
- '10.10.88.75:9100'
38+
- '10.10.88.76:9100'
39+
scrape_interval: 30s
2640

2741
alertmanager:
2842
alertmanagerSpec:
@@ -37,6 +51,9 @@ alertmanager:
3751

3852
grafana:
3953
adminPassword: "{{ grafana_admin_password | default('admin') }}"
54+
service:
55+
type: LoadBalancer
56+
loadBalancerIP: "{{ grafana_lb_ip | default('10.10.88.86') }}"
4057
persistence:
4158
enabled: true
4259
storageClassName: longhorn
@@ -45,22 +62,11 @@ grafana:
4562
dashboards:
4663
enabled: true
4764
searchNamespace: ALL
48-
dashboardProviders:
49-
dashboardproviders.yaml:
50-
apiVersion: 1
51-
providers:
52-
- name: 'default'
53-
orgId: 1
54-
folder: ''
55-
type: file
56-
disableDeletion: false
57-
editable: true
58-
options:
59-
path: /var/lib/grafana/dashboards/default
6065

61-
# Node exporter for host metrics
66+
# Node exporter disabled — npu-device-monitor DaemonSet already provides
67+
# node_exporter metrics on hostPort 9100 on all worker nodes
6268
nodeExporter:
63-
enabled: true
69+
enabled: false
6470

6571
# Kube-state-metrics
6672
kubeStateMetrics:

0 commit comments

Comments
 (0)