Add Horizon-equivalent dashboard for BackgroundJobs (throughput, retries, failures)

[antosubash]

Background

BackgroundJobs runs jobs but provides no operator UI. There is no way to inspect queue depth, throughput, failed-job payloads, or to retry/delete jobs without poking the DB. Laravel Horizon is the equivalent for Redis-queues.

Motivation

• Ops needs visibility into the queue without SSH+SQL
• Failed-job triage (view exception, edit payload, retry)
• Capacity planning (throughput per queue, per job-type)

Design sketch

Add an Inertia UI section under Admin (/admin/jobs) — not a separate module since it leans on BackgroundJobs:

Pages:

• Overview: queues, depth, throughput last 5 min / 1 h / 24 h, success vs failure rate
• Pending: list of queued jobs with payload preview, scheduled-at
• Running: in-flight jobs + duration
• Failed: stack trace, payload, last-attempt-at; actions: retry / retry-all / delete
• Completed: rolling 24 h window
• Recurring: scheduled-job definitions (depends on the Scheduler issue)

Backend:

• New IJobMetrics service with RecordEnqueued/Started/Completed/Failed hooks called by the runner
• Metrics persisted as 1-minute buckets per (queue, job-type, outcome) for cheap chart aggregation
• Failed jobs already persisted; add RetryAsync(jobId) API
• Auth: Jobs.View and Jobs.Manage permissions

Acceptance criteria

• IJobMetrics + bucket persistence
• Inertia pages with Recharts (or current chart lib) charts
• Retry / delete / bulk-retry actions
• Permission-gated and tested
• xUnit + integration tests
• NBomber load test for /admin/jobs/overview
• Docs page

References

• Laravel Horizon: https://laravel.com/docs/horizon
• Existing module: modules/BackgroundJobs