GCP.Monitoring reference

AlertPolicy

Source: src/GCP/Monitoring/AlertPolicy.ts

A Cloud Monitoring alerting policy — conditions that open incidents and optional notification channels.

Policy ids are assigned by the API. Alchemy stamps ownership into userLabels so list / pnpm nuke:gcp can find them. Display name, combiner, conditions, channels, documentation, enabled, severity, strategy, and labels update in place.

AlertPolicy: Creating a Policy

CPU threshold without a notification channel

const policy = yield* GCP.Monitoring.AlertPolicy("CpuHigh", {
  combiner: "OR",
  conditions: [
    {
      displayName: "CPU > 90%",
      conditionThreshold: {
        filter:
          'resource.type = "gce_instance" AND metric.type = "compute.googleapis.com/instance/cpu/utilization"',
        comparison: "COMPARISON_GT",
        thresholdValue: 0.9,
        duration: "60s",
        aggregations: [
          { alignmentPeriod: "60s", perSeriesAligner: "ALIGN_MEAN" },
        ],
      },
    },
  ],
});

Policy that emails a channel

const channel = yield* GCP.Monitoring.NotificationChannel("Oncall", {
  type: "email",
  labels: { email_address: "oncall@example.com" },
});
const policy = yield* GCP.Monitoring.AlertPolicy("CpuHigh", {
  conditions: [
    {
      displayName: "CPU > 90%",
      conditionThreshold: {
        filter:
          'resource.type = "gce_instance" AND metric.type = "compute.googleapis.com/instance/cpu/utilization"',
        comparison: "COMPARISON_GT",
        thresholdValue: 0.9,
        duration: "60s",
      },
    },
  ],
  notificationChannels: [channel.name],
});

AlertPolicy: Updating a Policy

const policy = yield* GCP.Monitoring.AlertPolicy("CpuHigh", {
  combiner: "OR",
  enabled: false,
  conditions: [
    {
      displayName: "CPU > 50%",
      conditionThreshold: {
        filter:
          'resource.type = "gce_instance" AND metric.type = "compute.googleapis.com/instance/cpu/utilization"',
        comparison: "COMPARISON_GT",
        thresholdValue: 0.5,
        duration: "60s",
      },
    },
  ],
});

Group

Source: src/GCP/Monitoring/Group.ts

A Cloud Monitoring group — a dynamic collection of monitored resources selected by a filter.

Group ids are assigned by the API. Groups have no labels field, so Alchemy stamps ownership into displayName ([alchemy alchemy-stack=… alchemy-stage=… alchemy-id=…]) so list / pnpm nuke:gcp can find them. Display name, filter, parent, and cluster flag update in place.

Group: Creating a Group

Root group of GCE instances in a region

const group = yield* GCP.Monitoring.Group("Prod", {
  displayName: "production instances",
  filter: 'resource.metadata.region="us-central1"',
});

Nested group

const parent = yield* GCP.Monitoring.Group("Prod", {
  filter: 'resource.metadata.region="us-central1"',
});
const child = yield* GCP.Monitoring.Group("Workers", {
  filter: 'resource.metadata.tag="worker"',
  parentName: parent.name,
});

Group: Updating a Group

const group = yield* GCP.Monitoring.Group("Prod", {
  displayName: "production instances",
  filter: 'resource.metadata.region="us-east1"',
  isCluster: true,
});

MetricDescriptor

Source: src/GCP/Monitoring/MetricDescriptor.ts

A Cloud Monitoring metric descriptor — the schema for a custom or external metric type.

Descriptors have no resource labels. Alchemy stamps ownership into description ([alchemy alchemy-stack=… alchemy-stage=… alchemy-id=…]) so list / pnpm nuke:gcp can find them. type, metricKind, and valueType are immutable — changing them replaces the descriptor. Display name, description, unit, metadata, and added labels update in place via create-as-upsert.

MetricDescriptor: Creating a Descriptor

Generated custom gauge

const metric = yield* GCP.Monitoring.MetricDescriptor("Paid", {
  displayName: "Invoice paid amount",
  description: "amount collected per invoice",
});

Explicit type, kind, and labels

const metric = yield* GCP.Monitoring.MetricDescriptor("Paid", {
  type: "custom.googleapis.com/invoice/paid/amount",
  metricKind: "DELTA",
  valueType: "INT64",
  unit: "1",
  labels: [{ key: "currency", valueType: "STRING" }],
});

MetricDescriptor: Updating a Descriptor

const metric = yield* GCP.Monitoring.MetricDescriptor("Paid", {
  displayName: "Invoice paid (cents)",
  unit: "1",
  description: "amount collected per invoice in cents",
});

NotificationChannel

Source: src/GCP/Monitoring/NotificationChannel.ts

A Cloud Monitoring notification channel — email, webhook, Pub/Sub, or another supported delivery type.

Channel ids are assigned by the API. Alchemy stamps ownership into userLabels so list / pnpm nuke:gcp can find them. Changing type replaces the channel. Display name, description, labels, user labels, and enabled update in place.

NotificationChannel: Creating a Channel

Email channel

const channel = yield* GCP.Monitoring.NotificationChannel("Alerts", {
  type: "email",
  labels: { email_address: "alerts@example.com" },
});

Webhook channel with user labels

const channel = yield* GCP.Monitoring.NotificationChannel("Hooks", {
  type: "webhook_tokenauth",
  displayName: "incident webhook",
  labels: { url: "https://example.com/hooks/alerts" },
  userLabels: { env: "prod" },
});

NotificationChannel: Updating a Channel

const channel = yield* GCP.Monitoring.NotificationChannel("Alerts", {
  type: "email",
  labels: { email_address: "alerts@example.com" },
  enabled: false,
});

Service

Source: src/GCP/Monitoring/Service.ts

A Cloud Monitoring service — the root resource for SLO monitoring.

Custom services (custom: {}) are the default when no other identifier is set. Alchemy stamps ownership into userLabels so list / pnpm nuke:gcp can find them. Changing serviceId or the identifier kind replaces the service. Display name, labels, and telemetry update in place.

Service: Creating a Service

Generated custom service

const checkout = yield* GCP.Monitoring.Service("Checkout", {
  displayName: "Checkout",
  labels: { env: "prod" },
});

Explicit id

const checkout = yield* GCP.Monitoring.Service("Checkout", {
  serviceId: "checkout",
  displayName: "Checkout API",
});

Service: Updating a Service

const checkout = yield* GCP.Monitoring.Service("Checkout", {
  displayName: "Checkout v2",
  labels: { env: "prod", team: "payments" },
});

ServicesServiceLevelObjective

Source: src/GCP/Monitoring/ServicesServiceLevelObjective.ts

A Cloud Monitoring service-level objective under a Monitoring Service.

Alchemy stamps ownership into userLabels so list / pnpm nuke:gcp can find them. Changing service or serviceLevelObjectiveId replaces the SLO. Goal, period, SLI, display name, and labels update in place.

ServicesServiceLevelObjective: Creating an SLO

const checkout = yield* GCP.Monitoring.Service("Checkout", {});
const slo = yield* GCP.Monitoring.ServicesServiceLevelObjective(
  "Latency",
  {
    service: checkout.name,
    goal: 0.99,
    rollingPeriod: "86400s",
    serviceLevelIndicator: {
      requestBased: {
        distributionCut: {
          distributionFilter:
            'metric.type="serviceruntime.googleapis.com/api/request_latencies" AND resource.type="consumed_api"',
          range: { max: 500 },
        },
      },
    },
  },
);

ServicesServiceLevelObjective: Updating an SLO

const slo = yield* GCP.Monitoring.ServicesServiceLevelObjective(
  "Latency",
  {
    service: checkout.name,
    goal: 0.95,
    calendarPeriod: "MONTH",
    serviceLevelIndicator: {
      requestBased: {
        distributionCut: {
          distributionFilter:
            'metric.type="serviceruntime.googleapis.com/api/request_latencies" AND resource.type="consumed_api"',
          range: { max: 500 },
        },
      },
    },
  },
);

UptimeCheckConfig

Source: src/GCP/Monitoring/UptimeCheckConfig.ts

A Cloud Monitoring uptime check — an HTTP, HTTPS, TCP, or synthetic probe that measures availability.

Config ids are assigned by the API. Alchemy stamps ownership into userLabels so list / pnpm nuke:gcp can find them. Display name, period, timeout, HTTP/TCP settings, matchers, regions, and labels update in place.

UptimeCheckConfig: Creating a Check

HTTPS probe of example.com

const check = yield* GCP.Monitoring.UptimeCheckConfig("Homepage", {
  httpCheck: { path: "/", useSsl: true, validateSsl: true },
  monitoredResource: {
    type: "uptime_url",
    labels: { host: "example.com" },
  },
});

Generated name with labels

const check = yield* GCP.Monitoring.UptimeCheckConfig("Homepage", {
  timeout: "10s",
  period: "600s",
  labels: { env: "prod" },
});

UptimeCheckConfig: Updating a Check

const check = yield* GCP.Monitoring.UptimeCheckConfig("Homepage", {
  period: "300s",
  httpCheck: { path: "/health", useSsl: true, validateSsl: true },
  monitoredResource: {
    type: "uptime_url",
    labels: { host: "example.com" },
  },
});