将 Azure 监控警报设置为 100% CPU 30 分钟

Question

所以我定义了以下 Terraform 资源。我不太确定窗口大小到底是多少。 % 连续 30 分钟？

resource "azurerm_monitor_metric_alert" "cpu_percent" {
  name                = "alert-${azurerm_postgresql_flexible_server.pgsql_server.name}-cpupercent"
  resource_group_name = data.azurerm_resource_group.resource_group.name
  scopes              = [azurerm_postgresql_flexible_server.pgsql_server.id]
  description         = "Action will be triggered Whenever the maximum cpu percent is greater than 100% for 30 minutes."
  window_size         = "PT30M"
  frequency           = "PT15M"
  severity            = 0

  criteria {
    metric_namespace = "Microsoft.DBforPostgreSQL/flexibleServers"
    metric_name      = "cpu_percent"
    aggregation      = "Maximum"
    operator         = "Equals"
    threshold        = 100
  }

我是否正确地假设这意味着 Azure 每 15 分钟检查一次数据库，并在最大 CPU 为 100 时发出警报？

Answer 1

使用 terraform 设置 Azure 监控警报 100% CPU 30 分钟

Terraform 配置，为 Azure PostgreSQL 灵活服务器上的 CPU 使用情况设置指标警报。如果 CPU 使用率在过去 30 分钟内的任何时候达到 100%，此警报应每 15 分钟检查一次，就像我在下面提到的那样。

我的 Terraform 配置：

provider "azurerm" {
  features {}
}

resource "azurerm_resource_group" "resource_group" {
  name     = "vksb-resources"
  location = "East US"
}

resource "azurerm_postgresql_flexible_server" "pgsql_server" {
  name                = "vksbepgserver"
  resource_group_name = azurerm_resource_group.resource_group.name
  location            = azurerm_resource_group.resource_group.location
  sku_name            = "GP_Standard_D4s_v3"
  administrator_login = "adminuser"
  administrator_password = "H@Sh1CoR3!"
  storage_mb = 32768
  
  version = "12"
}

data "azurerm_resource_group" "resource_group" {
  name = azurerm_resource_group.resource_group.name
}

resource "azurerm_monitor_metric_alert" "cpu_percent" {
  name                = "alert-${azurerm_postgresql_flexible_server.pgsql_server.name}-cpupercent"
  resource_group_name = data.azurerm_resource_group.resource_group.name
  scopes              = [azurerm_postgresql_flexible_server.pgsql_server.id]
  description         = "Action will be triggered whenever the maximum CPU percent is greater than 100% for 30 minutes."
  window_size         = "PT30M"
  frequency           = "PT15M"
  severity            = 0

  criteria {
    metric_namespace = "Microsoft.DBforPostgreSQL/flexibleServers"
    metric_name      = "cpu_percent"
    aggregation      = "Maximum"
    operator         = "Equals"
    threshold        = 100
  }
}

部署成功：