Ceph » mgr » Dashboard

Hosts should have actions (Edit and Delete), but

• Delete only deletes from the UI, and has no effect on daemons
• Edit presents a modal for label management, but existing labels are not shown and there is no information at all to guide the user.
• Maintenance is missing
• Drain is missing (part of host removal workflow)
• Refresh is missing - following a host upgrade, you would expect to see changes in metadata, so a manual way to refresh should be provided (just invoke gather-facts)

• No metadata is available for the host itself (model, make, bios, hdds, flash) etc (see below)
• The device health shows Unknown since it requires smartctl instead of using data from ceph-volume (coming soon!)
• Performance details shows anomalies with OSD capacity, as N/A and performance charts with errors (duplicate series)
• Daemons tab has a Last seen column with an overly verbose timestamp (e.g. 2020-08-05T02:31:43.066390)
• Daemons are systemd services, but there is no way from the UI of managing those daemons - which forces the admin back to the CLI!
  • Restart is missing
  • Stop is missing
  • Start is missing
  • Remove is missing
  • Fetch logs (provide a means of gathering the logs for support case resolution)
• Inventory will be covered separately, since this is an embedded component
• Device Health doesn’t seem to work. It reports a “Unable to retrieve smart data” failure in the UI - but there isn’t a corresponding notification!. (If ceph-volume reports device health, this tab should could be removed as the UI pivots to be more user friendly)

Server info mock-up: