gcp-status

GCP Status Exporter

Converts Google Cloud Status Dashboard into a series of up metrics (by services) for Prometheus consumption.

# HELP gcp_status_build_info A metric with a constant '1' value labeled by OS version, Go version, and the Git commit of the exporter
# TYPE gcp_status_build_info counter
gcp_status_build_info{git_commit="",go_version="go1.18",os_version=""} 1
# HELP gcp_status_services Count of GCP service availability
# TYPE gcp_status_services gauge
gcp_status_services{region="Americas"} 94
gcp_status_services{region="Asia Pacific"} 85
gcp_status_services{region="Europe"} 95
gcp_status_services{region="Global"} 61
gcp_status_services{region="Multi-regions"} 21
# HELP gcp_status_services_total Count of GCP services
# TYPE gcp_status_services_total gauge
gcp_status_services_total 138
# HELP gcp_status_start_time Exporter start time in Unix epoch seconds
# TYPE gcp_status_start_time gauge
gcp_status_start_time 1.659727534e+09
# HELP gcp_status_up Status of GCP service (1=Available; 0=Unavailable)
# TYPE gcp_status_up gauge
gcp_status_up{region="Americas",service="Artifact Registry"} 1
gcp_status_up{region="Asia Pacific",service="Artifact Registry"} 1
gcp_status_up{region="Europe",service="Artifact Registry"} 1
gcp_status_up{region="Multi-regions",service="Artifact Registry"} 1
gcp_status_up{region="Americas",service="Google Kubernetes Engine"} 1
gcp_status_up{region="Asia Pacific",service="Google Kubernetes Engine"} 1
gcp_status_up{region="Europe",service="Google Kubernetes Engine"} 1

Metrics

All metric names are prefix gcp_status_

Rules

If the listed Americas-based services are down for 15 minutes as measured in 5-minute chunks

- alert: gcp_status_up_down_americas
  expr: |
    min_over_time(
      gcp_status_up{
        region="Americas",
        service=~"Artifact Registry|Google Compute Engine|Google Kubernetes Engine|Cloud Firestore|Google Cloud Functions|Cloud Run|Google Cloud Scheduler"
      }[5m])<1
  for: 15m
  labels:
    severity: page
  annotations:
    summary: GCP Status service {{ $labels.service }} down in {{ $labels.region }}

If the listed Global services are down for 15 minutes as measured in 5-minute chunks

- alert: gcp_status_up_down_global
  expr: |
    min_over_time(
      gcp_status_up{
        region="Global",
        service=~"Cloud Endpoints|Cloud Logging|Cloud Monitoring|Cloud Profiler"
        }[5m])<1
  for: 15m
  labels:
    severity: page
  annotations:
    summary: GCP Status service {{ $labels.service }} down in {{ $labels.region }}

Run

Go

go run .

Docker

docker run \
--interactive --tty --rm \
ghcr.io/dazwilkin/gcp-status:ae3fc74f3245a46d984c040b863e2a2980ff3ce2 \
--endpoint=:9989 \
--path=/metrics

Docker Compose

gcp-exporter:
  image: ghcr.io/dazwilkin/gcp-status:ae3fc74f3245a46d984c040b863e2a2980ff3ce2
  container_name: gcp-status
  expose:
  - "9989" # GCP Status port registered on Prometheus Wiki
  ports:
  - 9989:9989