Skip to content
shadcn.io

shadcn.io is not affiliated with official shadcn/ui

Status Ops Board Features Block for React

An error-budget board for an SLO monitor: five services in bands with a budget bar, a 24-point burn sparkline and a status each, a 1h, 6h, 24h and 30d toggle, and a summary counted from the rows.

Framework
UI library

Preview

Preview width

Installation

Part of Pro
Pro unlocks the code and one-line install of every Pro button and text effect, plus the MCP for your AI editor.
Have Pro already? Sign in to open it.Opens here as soon as you're Pro.

Overview

An SLO error-budget board for React with Base UI. A left-aligned header says "Know which budget is burning" and holds a four-button toggle at its end (1h, 6h, 24h, 30d). Under it, bands that run from rail to rail, each ruled by a hairline with + marks on the rails, hold five services: its name and target, a bar for the share of its error budget that is left with the figure rolling beside it, a 24-point burn sparkline with its burn rate, and a status with a dot. Pressing a toggle button swaps every row to its reading for that span, and the footer band's summary ("3 healthy, 1 burning, 1 exhausted") is counted from the rows on screen. On a phone each row is a card of three lines. It's built from Guides, Progress, Status, Toggle Group and Sliding Number, with an SVG sparkline.

FAQ

How do I use my own services?

Edit SERVICES: each has a name, a target and one Reading per span, which is the share of the error budget left (0 when it is spent) and the shape of the burn (flat, rising, falling or spike). The sparkline is drawn from the shape plus a little noise seeded by the service, so it is the same on the server and in the browser; to draw a real series, return it from sparkOf as 24 numbers between 0 and 1. The conditions, the burn rates and the footer's summary are all counted from the readings.

When is a service burning or exhausted?

Under 20% of the budget left, a service is burning, and its bar takes the warning tone. With none left it is exhausted. Otherwise it is healthy. The status is also written as a word, so colour is never the only signal, and the threshold is BURNING_UNDER at the top of the file.

What is the burn rate?

It is the average burn over the last six points of the sparkline, where 1× spends the budget exactly at the pace the target allows and 2× spends it twice as fast. It is counted by rateOf from the sparkline's own points, so the number and the line agree.

How is the board made accessible?

Each row is a list item with the service as a heading, and its figures are text: the target, the share of the budget left (with an unseen "of the error budget left"), the burn rate and the status word. The bars and sparklines are decoration. The footer's summary is a polite live region, so pressing a toggle button announces the new counts.

How does it fit on a phone?

The columns become three lines inside each row: the name with its status, the budget bar with its figure, then the sparkline with its rate. The column headings are dropped. Nothing scrolls sideways at 320 pixels.

What happens in a right-to-left layout?

The bars and the sparklines keep their direction in every language, because a time series does. Names, targets, statuses and the footer follow the page, and values with a unit are wrapped so they don't reorder.

  • Pro

    A metrics console in a card in a card that follows your theme: six metric tiles with a value, a signed change and a sparkline are the tabs of one tablist, and choosing one opens a larger chart of its series in a strip below.

  • Pro

    A split for a platform SLA monitor: copy, four ticks and a button at the start, and at the end one chart of three concentric rings with a legend, whose values a 7, 30 or 90 day toggle swaps.

  • Pro

    A security-posture scanner: eight checks in a hairline grid with a beam that sweeps across it, a Pause scan toggle, and the chosen check's finding, fix and last scan in a band under it.

  • Pro

    Gauge Cluster Features

    Metrics and charts

    Three semicircle gauges in one framed band, driven by a single "Simulate load" slider: CPU, memory and queue wait, each with a needle, an arc, a reading and a state.

  • Pro

    Usage Meters Billing

    Plan and usage

    Usage meters for an app's settings: one row per quota with its count, a bar that turns to a warning tone at 80% and an over-limit tone past 100%, and a plain line on what happens next.

  • Pro

    A showcase centered hero for an API platform, with a live status pill, a Shine Button and a status panel of full-width bands: five services' uptime over the last 30 days.