Custom Dashboards - Your Metrics, Your Layout, Zero Config Files

The template gallery ships 18 ready-made dashboards. Search it, filter by category, and click once - the dashboard is created with every panel already sized and placed.

Before it builds, you choose which entities it applies to. A fleet of 50 servers does not have to mean 50 lines on every chart: pick the servers, uptime monitors, tasks, network devices or Ceph clusters you actually want, and rename the dashboard while you are there.

Templates only create panels that will have data. If a template covers RAID, S.M.A.R.T. and ZFS but you only run ZFS, the panels that would read "No data" are dropped and you are told how many and why.

Already curated a dashboard you like? Add a template to it as a new collapsible section instead of creating a second dashboard. Everything you already built stays exactly where it is.

Panels drop onto a 24-column grid, so two, three or four across all line up cleanly. Drag them between groups, resize them, and let the layout save itself.

Group related panels into collapsible sections - a Grafana-style row - and collapse the ones you are not looking at today. A collapsed section does not run its queries until you open it, so a long dashboard stays fast.

Turn on auto-refresh and every panel re-reads every five seconds. Hit fullscreen for the wall display. Hover one chart and the matching timestamp lights up on the others, so you can line up a spike across panels without guessing.

Zoom into a five-minute window to investigate an incident, or zoom out across months of history - up to your plan's retention window, which runs from six months to two years. Export a time-series chart as a PNG when you need it in a report or a ticket.

Panels that read your incident record - availability, incident count, MTTR and MTTA - come from the same incidents your alerts opened, so the dashboard and the incident list can never disagree.