Azure 检查
azure 模块有 10 项检查。通过 Resource Graph (KQL) 和 Azure Monitor 指标检查 Azure 平台,使用采集器 VM 的托管标识(Reader + Monitoring Reader)。
本页在构建文档时根据产品的检查目录生成,因此规则、间隔和默认严重性始终与你正在阅读的版本一致。阈值是默认值;实体和资源可以覆盖这些阈值。
| 检查 | 间隔 | 默认严重性 | 规则 |
|---|---|---|---|
azure.vm_power_stateVM power state | 120s | critical | power = PowerState/running |
azure.vm_cpuVM CPU (host view) | 300s | warning | warning > 80 percent · 15m / critical > 90 percent · 5m |
azure.sqldb_dtuAzure SQL Database DTU / vCore | 300s | warning | warning > 80 percent · 15m / critical > 95 percent · 5m |
azure.sqldb_storageAzure SQL Database storage used | 900s | warning | warning > 80 percent / critical > 90 percent |
azure.sqldb_deadlocksAzure SQL Database deadlocks | 300s | warning | warning > 0 count / critical >= 5 count |
azure.pgflex_healthPostgreSQL Flexible Server CPU, memory, storage | 300s | warning | warning > 80 percent · 15m / critical > 90 percent · 5m |
azure.vpn_tunnelsVPN gateway tunnel state | 120s | critical | connection_status = Connected |
azure.appgw_backend_healthApplication Gateway backend health and 5xx | 120s | warning | warning >= 1 count / critical >= 2 count |
azure.backup_jobsRecovery Services Vault backup jobs | 3600s | critical | warning >= 1 count / critical >= 1 count |
azure.service_healthAzure Service Health events | 600s | info | warning >= 1 count |
检查
azure.vm_power_state
VM power state
- 度量
power_state· 每120s·azure_graph- 规则
power = PowerState/running· 默认严重性: critical- 措施
- —
azure.vm_cpu
VM CPU (host view)
- 度量
cpu_percent(percent) · 每300s·azure_monitor- 规则
warning > 80 percent · 15m / critical > 90 percent · 5m· 默认严重性: warning- 措施
- —
azure.sqldb_dtu
Azure SQL Database DTU / vCore
- 度量
dtu_percent(percent) · 每300s·azure_monitor- 规则
warning > 80 percent · 15m / critical > 95 percent · 5m· 默认严重性: warning- 措施
- —
azure.sqldb_storage
Azure SQL Database storage used
- 度量
storage_percent(percent) · 每900s·azure_monitor- 规则
warning > 80 percent / critical > 90 percent· 默认严重性: warning- 措施
- —
azure.sqldb_deadlocks
Azure SQL Database deadlocks
- 度量
deadlocks(count) · 每300s·azure_monitor- 规则
warning > 0 count / critical >= 5 count· 默认严重性: warning- 措施
- —
azure.pgflex_health
PostgreSQL Flexible Server CPU, memory, storage
- 度量
cpu_percent(percent) · 每300s·azure_monitor- 规则
warning > 80 percent · 15m / critical > 90 percent · 5m· 默认严重性: warning- 措施
- —
- 基线
- Storage uses 85 / 95 as override defaults in the normalizer.
azure.vpn_tunnels
VPN gateway tunnel state
- 度量
tunnel_up· 每120s·azure_monitor- 规则
connection_status = Connected· 默认严重性: critical- 措施
- —
- 基线
- Also read connectionStatus from Resource Graph for each connection.
azure.appgw_backend_health
Application Gateway backend health and 5xx
- 度量
unhealthy_hosts(count) · 每120s·azure_monitor- 规则
warning >= 1 count / critical >= 2 count· 默认严重性: warning- 措施
- —
azure.backup_jobs
Recovery Services Vault backup jobs
- 度量
failed_jobs(count) · 每3600s·azure_graph- 规则
warning >= 1 count / critical >= 1 count· 默认严重性: critical- 措施
- —
- 基线
- Any failed backup job in the last 48 h is critical for production items.
azure.service_health
Azure Service Health events
- 度量
active_events(count) · 每600s·azure_graph- 规则
warning >= 1 count· 默认严重性: info- 措施
- —
- 基线
- PlannedMaintenance → advisory; ServiceIssue → warning on affected regions.