Dashboard UI — User Guide

This guide covers the Playwright Grid Dashboard for viewing runs, results, and live logs.

Overview - Live run feed via SignalR (auto-reconnect with backoff). - Runs list with filtering, sorting, and paging. - Run details: summary, timeline, per-browser command logs, and metadata. - Deep links to runs by runId.

Access - Default URL in local compose: http://127.0.0.1:3001 - Configure HUB_SIGNALR for the Dashboard to point to the Hub’s ws endpoint (see docker-compose.yml).

Runs page - Columns: RunId, App, Browser, Env, Status, Started, Completed, Duration, Failures. - Filters: App, Browser, Env, Region, Status, runId. Use the filter bar to narrow down. - Pagination and virtualization keep the UI responsive on large datasets.

Run details - Summary header shows core metadata and status. - Command log panel streams events (worker protocol and optional runner logs) with time ordering. - Use the search box to filter commands by keyword. - Copy runId using the action in the header for sharing.

Live feed and resilience - The Dashboard uses SignalR to receive updates. If disconnected, it shows an inline warning and automatically retries with exponential backoff. - Error boundaries provide friendly messages and retry actions when transient errors occur.

Troubleshooting - No runs appearing: - Ensure the Hub is reachable from the Dashboard (HUB_SIGNALR env). - Verify Hub /health is OK and Redis is up. - Logs not streaming: - Check browser console for SignalR connection errors (CORS, URL mismatch). - Confirm the runId is correct and that workers are forwarding protocol logs. - Missing data or slow UI on large runs: - The Results pages use server-driven paging and streaming; verify Hub and Redis health.

Related docs - Overview & Quick Start: docs-guide.md - Metrics and Grafana: Metrics-and-Grafana.md - Capacity Queue: Capacity-Queue.md - Label Matching: Label-Matching.md