Queue & scheduling

The Queue page is the central view of what the bot is doing right now and what’s pending. Everything goes through Redis (queue + history + state), so this view is restart-safe and reflects reality across all worker processes — including a worker that just crashed and restarted.

What you get

Per-instance filter

Pick one BlueStacks instance (bs1, 127.0.0.1:5555, …) or view the whole fleet. The page subscribes to live Redis updates — no manual refresh.

At-a-glance counters

Pending, Overdue, Running, Recent OK, Recent failed. The first place to look when something feels off.

Currently running task

What’s executing on which instance, with priority, current step, elapsed time, and the player ID. One click opens the DSL runner with the same state.

Pending table

Sorted per instance in execution order — due tasks first, then scheduled by time. Each row shows status, scenario, target region, priority, and quick actions.

Two views: Table and Timeline

The Table view is a list — best for “what’s stuck right now.”

The Timeline view is a Google-Calendar-style grid of upcoming scenarios across your instances. Each block is one scheduled task; rows are instances; columns are time. You can:

• See at a glance when each daily/weekly scenario is going to fire next.
• Spot scheduling conflicts (two heavy scenarios stacked on one minute → throttle one).
• Drag a task to reschedule it to a later slot — the change writes straight to Redis and the scheduler picks it up.

This makes the cron schedule a UI-editable object, not a YAML file you have to redeploy.

Cross-references

The Queue page is wired to every other operator surface — the top-right action row jumps to:

• Labeling — open the labeling editor for the regions the running task is matching against
• Player — open the active player’s profile
• Approvals — review pending tap approvals
• Overlay — see what the overlay engine is detecting right now

Use it as the “everything’s wrong, where do I start” entry point.