Skip to content
shadcn.io

shadcn.io is not affiliated with official shadcn/ui

Pipeline Run Task Duration Monitoring Block for Solid

The last eight runs of a pipeline as stacked columns of five tasks, a failed run hatched at its failing task, with the chosen run's DAG drawn as boxes, a text route and each task's change against its median.

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

Say which task made a run slow, not just that it was slow, with this run history for Solid with Ark UI. It is Dagmark, a workflow monitor following the nightly_warehouse_load DAG on Airflow. The last eight runs are stacked columns, each stack the five tasks in order (extract, validate, transform, load, notify) in five weights, with a failed run hatched at the task that failed and nothing above it. The columns are one radio group, so arrow keys move between runs. Choosing a run draws the DAG as five boxes joined by arrows, each with its state, and a text route under them ("Run 4 stopped at transform"), then fills a table with each task's duration and how far it is over or under the median of the earlier runs that finished. The header sentence names the slowest run and the task that added the most. Guides draw the frame and its bands. For data teams, nightly loads and SLA reviews.

FAQ

How do I load my own runs?

Edit TASKS and RUNS: each run has a number, a date, the minutes each task took (0 for one that did not run) and the index of the task that failed, or -1. Medians, the slowest run and the culprit task are derived from those minutes.

What is the median compared with?

The same task in the earlier runs that finished, so a failed run never pulls the median down. The first run has no earlier runs and says so.

Is anything measured?

No. Runs and minutes are typed examples and nothing calls a network or a workflow tool. Choosing a run only reads it.

What is the text equivalent of the chart?

Each column has a spoken name (run, date, total, minutes over the median or where it failed), the DAG boxes and the route sentence say the structure in words, and the table lists every task with its duration. The hatch and the weights are not the only signal.

How does it look on a phone?

The eight columns share the width, the arrows between boxes hide so the boxes wrap, and the table drops its median column, leaving the task, its time and its change.