All docs
Integrations & API
Custom dashboards and sharing
Dashboards let you put the numbers you care about on one page: a health gauge, open alerts, CPU over the last week, disk per mount as a table. Drag tiles into place, pick fleet-wide or specific servers, and share the page by link with a password if you need one.
Create a dashboard
Open Dashboards from the top navigation. Type a name, then either Create empty or Start from Fleet overview. The starter gives you eight tiles: health score, open critical alerts, services down, days to the soonest SSL expiry, CPU and memory charts, and tables of disk per mount and pending security updates. Delete or change any of them.
Add and edit tiles
Click + Add tile, or a tile’s title to edit it. Each tile has:
- •Metric — search or browse by group (CPU, Memory, Disk, Network, Web, Services, Containers, Security, Updates, Databases, System, Health). Each entry says whether it has history, is latest-only, or is “per mount”, “per service”, “per container”, “per instance” or “per domain”.
- •Title — optional; the metric’s name is used otherwise.
- •Servers — All servers or Choose servers…. A fleet tile combines values the way the metric defines: summed (counts, bytes), averaged (percentages), worst (fullest disk, highest usage) or lowest (health score, SSL days). The option label tells you which.
- •Show as — Number, Gauge, Sparkline, Line chart, Bars or Table. Line, Sparkline and Bars need a metric with history. Per-item metrics are best as a Table, which ranks every server, mount, service or instance.
- •Time window — Last hour, 6 hours, 24 hours, 7 days or 30 days for charted tiles.
The grid has 12 columns. A tile can be 1–12 columns wide and 1–8 rows tall; a dashboard holds up to 40 tiles. Drag a tile by its title to move it and use the arrow buttons to resize. Changes save automatically. Data refreshes every 30 seconds.
Who can see and edit
- •Everyone in the workspace can open every dashboard.
- •The person who created a dashboard, and workspace owners and admins, can edit, share and delete it.
- •Creating, deleting and changing sharing are recorded in the audit log.
Share a dashboard by link
Click Share. Turning sharing on creates a public page at /d/<slug> on the Opservo dashboard that shows the dashboard read-only to anyone who has the link. No Opservo login is needed. The public page refreshes every 30 seconds and carries no navigation, so it works on a wall display.
- •Protect — set a password of at least 4 characters. Visitors see a password prompt before any number loads. Use Change or Remove password later.
- •Revoke link — the link stops working immediately. Sharing again creates a new slug, so old links stay dead.
- •Open public link — the editor shows the live link once sharing is on, so you can check what visitors will see.
Public pages are rate-limited to 60 requests per minute per visitor address. The public page shows only the tiles’ values; it does not expose server details beyond the display name or hostname you chose.
Where the numbers come from
Tiles read the same metric catalogue as the Prometheus endpoint and the Grafana data source, computed from each server’s latest heartbeat. Health score, security score and open-alert counts match the fleet grid. Scalar metrics are charted from stored history; per-item metrics show their latest value only.
Last updated 2026-10-01. Found a mistake? Email hello@getopservo.com.