Skip to main content
The UCloud Exporter chart deploys a Prometheus exporter for UCloud services, enabling monitoring of your UCloud infrastructure.

​
What It Monitors

This exporter collects metrics from UCloud services including:
  • UHost - Virtual machine instances
  • UDB - Managed database services
  • ULB - Load balancer metrics
  • UDisk - Cloud disk storage
  • UMem - Managed Redis/Memcache
  • Other UCloud products

​
Installation

​
Prerequisites

You need UCloud API credentials (PublicKey and PrivateKey).

​
Basic Installation

helm repo add douban https://douban.github.io/charts
helm install ucloud-exporter douban/ucloud-exporter

​
Authentication Configuration

Never commit UCloud credentials to version control. Always use Kubernetes secrets or external secret management solutions.

​
Using Environment Variables

Configure credentials securely: 
extraEnvs:
  - name: UCLOUD_PUBLIC_KEY
    valueFrom:
      secretKeyRef:
        name: ucloud-credentials
        key: public-key
  - name: UCLOUD_PRIVATE_KEY
    valueFrom:
      secretKeyRef:
        name: ucloud-credentials
        key: private-key
  - name: UCLOUD_REGION
    value: "cn-bj2"
  - name: UCLOUD_PROJECT_ID
    value: "your-project-id"
​
extraEnvs
array
default:"[]"
Environment variables for credentials and configuration

​
Create Credentials Secret

kubectl create secret generic ucloud-credentials \
  --from-literal=public-key="YOUR_PUBLIC_KEY" \
  --from-literal=private-key="YOUR_PRIVATE_KEY"

​
Configuration

​
args
array
default:"[]"
Additional command-line arguments for the exporter
​
image.repository
string
default:"ghcr.io/douban/ucloud-exporter"
Container image repository
​
image.tag
string
default:"main"
Container image tag
​
containerPort
number
default:"9200"
Port the exporter listens on
​
service.port
number
default:"9200"
Service port to expose metrics

​
ServiceMonitor Configuration

For automatic Prometheus scraping via the Prometheus Operator: 
serviceMonitor:
  enabled: true
  interval: 60s
  labels:
    prometheus: kube-prometheus
​
serviceMonitor.enabled
boolean
default:"false"
Enable ServiceMonitor resource creation
​
serviceMonitor.interval
string
default:"30s"
Scrape interval (60s recommended for cloud APIs)
​
serviceMonitor.timeout
string
default:"10s"
Scrape timeout

​
Manual Prometheus Configuration

If not using the Prometheus Operator: 
scrape_configs:
  - job_name: 'ucloud-exporter'
    static_configs:
      - targets: ['ucloud-exporter:9200']
    scrape_interval: 60s
    scrape_timeout: 30s

​
Example Values

replicaCount: 1

image:
  repository: ghcr.io/douban/ucloud-exporter
  tag: "main"
  pullPolicy: IfNotPresent

extraEnvs:
  - name: UCLOUD_PUBLIC_KEY
    valueFrom:
      secretKeyRef:
        name: ucloud-credentials
        key: public-key
  - name: UCLOUD_PRIVATE_KEY
    valueFrom:
      secretKeyRef:
        name: ucloud-credentials
        key: private-key
  - name: UCLOUD_REGION
    value: "cn-bj2"
  - name: UCLOUD_PROJECT_ID
    value: "org-abc123"

service:
  type: ClusterIP
  port: 9200

serviceMonitor:
  enabled: true
  interval: 60s
  timeout: 30s
  labels:
    release: prometheus

resources:
  limits:
    cpu: 200m
    memory: 256Mi
  requests:
    cpu: 100m
    memory: 128Mi

​
Metrics Exposed

Metrics are available at the /metrics endpoint:
  • ucloud_uhost_cpu_usage - UHost CPU usage percentage
  • ucloud_uhost_memory_usage - UHost memory usage
  • ucloud_uhost_disk_usage - UHost disk usage
  • ucloud_udb_connections - Database connections
  • ucloud_udb_qps - Database queries per second
  • ucloud_ulb_connections - Load balancer connections
  • ucloud_udisk_iops - Disk IOPS

​
Prometheus Rules

Example alerting rules: 
prometheusRule:
  enabled: true
  rules:
    - alert: UCloudHighCPU
      expr: ucloud_uhost_cpu_usage > 80
      for: 5m
      labels:
        severity: warning
      annotations:
        summary: High CPU usage on UHost
        description: "UHost {{ $labels.instance_id }} CPU usage is {{ $value }}%"
    
    - alert: UCloudDatabaseConnections
      expr: ucloud_udb_connections > 1000
      for: 5m
      labels:
        severity: warning
      annotations:
        summary: High database connections
        description: "UDB {{ $labels.db_id }} has {{ $value }} connections"

​
Use Cases

​
Infrastructure Monitoring

Monitor your UCloud infrastructure: 
# Average CPU usage across all UHosts
avg(ucloud_uhost_cpu_usage)

# Instances with high memory usage
ucloud_uhost_memory_usage > 80

​
Database Performance

Track database performance: 
# Database QPS
sum by (db_id) (ucloud_udb_qps)

# Connection pool usage
ucloud_udb_connections / ucloud_udb_max_connections

​
Security Best Practices

Follow these security guidelines:
  1. Use sub-accounts - Create dedicated API credentials with minimal permissions
  2. Read-only access - Grant only monitoring/describe permissions
  3. Rotate credentials - Change API keys regularly
  4. Use Kubernetes secrets - Never hardcode credentials in values files
  5. Network policies - Restrict network access to the exporter
  6. Audit logging - Enable API audit logs in UCloud console

​
Resources

Build docs developers (and LLMs) love

Get started for freeTalk to us