d4c8683e68
This patch consits of: * New ops interface grafana_dashboard based on the existing reactive interface * New ops http interface. * Dashboard registers grafana dashboards with grafana. These dashboards are taken from upstream ceph *1. Ideally the charm would pick these up from ceph packaging but they are currently not included. With the exception of host-details.json and hosts-overview.json the dashboards are unchanged. host-details.json and hosts-overview.json both needed changed to make the compatable with the metrics telegraf is sending. * alertmanager-service and prometheus relations using the http ops interface allowing there api endpoints to be registered with the dashboard. NOTE: ceph-mon has an existing relation with Prometheus ( ceph-mon:prometheus <-> prometheus:target ) but prometheus does not publish its api endpoint on that relation. So, the dashboard adds a prometheus:website <-> ceph-dashboard:prometheus relation. This allows the dashboard to set set-prometheus-api-host. *1 https://github.com/ceph/ceph/tree/master/monitoring/grafana/dashboards Change-Id: Ic4522cc601895c9a79489df985a6e81fa70fb9e5 |
||
|---|---|---|
| .. | ||
| bundles | ||
| __init__.py | ||
| tests.yaml | ||