Dashboards and alerts
Your Grafana has a Registry dashboard in its Registry folder. It reads only your tenant’s own data, and shows nothing until your tenant has a registry. See Grafana dashboards for the other built-in dashboards.
The Registry dashboard
Section titled “The Registry dashboard”| Row | Panels |
|---|---|
| Storage | Stored, Storage limit, Used of limit, Repositories, Images, Untagged images, and Storage over time against the limit |
| Contents | Images by age, Artifacts by type, Repositories |
| Activity | Pushes, pulls and deletes over time; Pulls in range; Pushes in range |
| Access | Access tokens, Login sessions, Tokens expiring in 7 days, Tokens by access, and a Cluster connections table |
| Alerts | Your firing registry alerts |
Used of limit turns orange at 80% and red at 95%, the thresholds of the two storage alerts.
How fresh the figures are
Section titled “How fresh the figures are”Dashboard figures can lag activity by up to 15 minutes. Refresh the time range if a recent push or deletion is not visible yet.
Alerts
Section titled “Alerts”The four registry alerts, RegistryStorageAlmostFull, RegistryStorageFull,
RegistryAccessTokenExpiring and RegistryConnectionNotReady, and when each
fires, are listed on the Registry page. The dashboard’s
Alerts row shows the ones firing now.
Each alert has the label registry="registry". RegistryConnectionNotReady
also has connected_cluster, set to the cluster whose connection is not
ready. None has a cluster label, so a route that
matches on cluster does not catch them.
Routing alerts to your receivers
Section titled “Routing alerts to your receivers”Send the registry’s alerts to a receiver of your own with a route that matches
registry="registry". In Terraform, as one of the routes of
kupe_alertmanager_routes:
{ matchers = ["registry=\"registry\""] receiver = "platform-team"}See Notifications for receivers, routes and the same change in the console.