Inventory management (INV) telemetry in Google Distributed Cloud (GDC) air-gapped appliance tracks
unauthorized container images, hardware and software inventory exports, ClamAV
bucket scans, and controller reconciliations across infrastructure clusters.
This document lists each metric name, data type, emission interval, and label
set.

## Metrics list

| Name and description || Type | Emit duration | Labels |
|---|---|---|---|---|
| `inv_unauthorized_image` |||||
| Indicates whether an unauthorized container image is detected in a Pod (`1` when any container, init container, or ephemeral container image is not from an authorized registry; `0` when all images are authorized). Triggers alert `INV-E0001` (runbook `INV-R0001`). || Gauge | 1 m | - **pod**: The name of the Pod using the unauthorized image. - **namespace**: The namespace of the Pod. |
| `kube_pod_controller_restarts_total` |||||
| Total number of container restarts per controller Pod (recording rule aggregated from `kube_pod_container_status_restarts_total`). Used on the `INV: Golden Signals` dashboard to monitor the controller restart rate in `inv-system`. || Counter | 1 m | - **namespace** : The namespace of the controller Pod (for example, `inv-system`). - **controller**: The controller prefix extracted from the Pod name. - **container**: The name of the container. - **zone_id**: The identifier of the zone. |
| `inv:export_hardware:last_success_ts` |||||
| Unix timestamp in seconds of the last successful run of the `inv-export-hw-backend` CronJob in `gpc-system` (recording rule based on `kube_cronjob_status_last_successful_time`). || Gauge | 1 m | - **namespace** : The namespace of the CronJob (`gpc-system`). - **cronjob** : The name of the CronJob (`inv-export-hw-backend`). |
| `inv:export_hardware:last_success_age` |||||
| Time elapsed in seconds since the last successful run of the `inv-export-hw-backend` hardware inventory export CronJob, or `-1` if it has never succeeded. Used on the `INV: Golden Signals` dashboard and runbook `INV-R0011`. || Gauge | 1 m |   |
| `inv:export_hardware:next_scheduled_age` |||||
| Time remaining in seconds until the next scheduled run of the `inv-export-hw-backend` CronJob, or `-1` if not scheduled. Used on the `INV: Golden Signals` dashboard. || Gauge | 1 m |   |
| `inv:export_hardware:good` |||||
| Freshness indicator timestamp for the `inv-export-hw-backend` CronJob, clamped to 25 hours after `inv:export_hardware:last_success_ts`. Used on the `INV: Golden Signals` dashboard (`rate(inv:export_hardware:good[5m]) >bool 0`) to indicate whether the export pipeline is fresh (`1`) or stale (`0`). || Gauge | 1 m |   |
| `inv:export_hardware:total` |||||
| Current evaluation Unix timestamp in seconds for the hardware inventory export recording rule group. || Gauge | 1 m |   |
| `inv:export_images:last_success_ts` |||||
| Unix timestamp in seconds of the last successful run of the `inv-export-img-backend` CronJob in `gpc-system` (recording rule based on `kube_cronjob_status_last_successful_time`). || Gauge | 1 m | - **namespace** : The namespace of the CronJob (`gpc-system`). - **cronjob** : The name of the CronJob (`inv-export-img-backend`). |
| `inv:export_images:last_success_age` |||||
| Time elapsed in seconds since the last successful run of the `inv-export-img-backend` software inventory export CronJob, or `-1` if it has never succeeded. Used on the `INV: Golden Signals` dashboard and runbook `INV-R0012`. || Gauge | 1 m |   |
| `inv:export_images:next_scheduled_age` |||||
| Time remaining in seconds until the next scheduled run of the `inv-export-img-backend` CronJob, or `-1` if not scheduled. Used on the `INV: Golden Signals` dashboard. || Gauge | 1 m |   |
| `inv:export_images:good` |||||
| Freshness indicator timestamp for the `inv-export-img-backend` CronJob, clamped to 25 hours after `inv:export_images:last_success_ts`. Used on the `INV: Golden Signals` dashboard (`rate(inv:export_images:good[5m]) >bool 0`) to indicate whether the export pipeline is fresh (`1`) or stale (`0`). || Gauge | 1 m |   |
| `inv:export_images:total` |||||
| Current evaluation Unix timestamp in seconds for the software inventory export recording rule group. || Gauge | 1 m |   |
| `inv:bucket_clamav:last_success_ts` |||||
| Unix timestamp in seconds of the last successful run of the `inv-bucket-clamav-backend` CronJob in `inv-system` (recording rule based on `kube_cronjob_status_last_successful_time`). || Gauge | 1 m | - **namespace** : The namespace of the CronJob (`inv-system`). - **cronjob** : The name of the CronJob (`inv-bucket-clamav-backend`). |
| `inv:bucket_clamav:last_success_age` |||||
| Time elapsed in seconds since the last successful run of the `inv-bucket-clamav-backend` ClamAV bucket scan CronJob, or `-1` if it has never succeeded. Used on the `INV: Golden Signals` dashboard. || Gauge | 1 m |   |
| `inv:bucket_clamav:next_scheduled_age` |||||
| Time remaining in seconds until the next scheduled run of the `inv-bucket-clamav-backend` CronJob, or `-1` if not scheduled. Used on the `INV: Golden Signals` dashboard. || Gauge | 1 m |   |
| `inv:bucket_clamav:good` |||||
| Freshness indicator timestamp for the `inv-bucket-clamav-backend` CronJob, clamped to 25 hours after `inv:bucket_clamav:last_success_ts`. Used on the `INV: Golden Signals` dashboard (`rate(inv:bucket_clamav:good[5m]) >bool 0`) to indicate whether the scan pipeline is fresh (`1`) or stale (`0`). || Gauge | 1 m |   |
| `inv:bucket_clamav:total` |||||
| Current evaluation Unix timestamp in seconds for the ClamAV bucket scan recording rule group. || Gauge | 1 m |   |
| `inv:audit_controller:bad` |||||
| Cumulative container restarts for the `inv-audit-backend` controller (recording rule filtering `kube_pod_controller_restarts_total{controller="inv-audit-backend"}`). || Counter | 1 m | - **namespace**: The namespace of the controller Pod. - **controller** : The controller name (`inv-audit-backend`). - **container**: The name of the container. - **zone_id**: The identifier of the zone. |
| `inv:audit_controller:total` |||||
| Current evaluation Unix timestamp in seconds for the `audit-controller` recording rule group. || Gauge | 1 m |   |
| `inv:expansion_controller:bad` |||||
| Cumulative container restarts for the `inv-expansion-backend` controller (recording rule filtering `kube_pod_controller_restarts_total{controller="inv-expansion-backend"}`). || Counter | 1 m | - **namespace**: The namespace of the controller Pod. - **controller** : The controller name (`inv-expansion-backend`). - **container**: The name of the container. - **zone_id**: The identifier of the zone. |
| `inv:expansion_controller:total` |||||
| Current evaluation Unix timestamp in seconds for the `expansion-controller` recording rule group. || Gauge | 1 m |   |
| `controller_runtime_reconcile_total` |||||
| Total number of reconciliations per controller exposed on port `8080` (`/metrics`) by the `inv-core`, `inv-audit`, and `inv-expansion` controller managers. || Counter | 5 m | - **controller**: The name of the reconciler controller. - **result** : The reconciliation outcome (`success`, `error`, `requeue`, or `requeue_after`). |
| `controller_runtime_reconcile_errors_total` |||||
| Total number of reconciliation errors per controller exposed by the `inv-core`, `inv-audit`, and `inv-expansion` controller managers. || Counter | 5 m | - **controller**: The name of the reconciler controller. |
| `controller_runtime_reconcile_time_seconds` |||||
| Duration in seconds of each reconciliation per controller exposed by the `inv-core`, `inv-audit`, and `inv-expansion` controller managers. || Histogram | 5 m | - **controller**: The name of the reconciler controller. |