> ## Documentation Index
> Fetch the complete documentation index at: https://docs.dishink.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Stuck Tables

> Diagnose and force-reset tables that show busy on the floor but have no session on the server.

A **Stuck** table is one where the floor plan shows a busy status (Ordering, Preparing, etc.) but the Table Session modal can't find an actual session. It's rare, and always fixable.

## Overview

Stuck happens when a table's status was written to the server but the matching session record didn't land — usually because of a network drop mid-write, or a partial sync from an offline device.

DishInk detects this state automatically and marks the card with a **Stuck** badge after a **600 ms delay** (to avoid flashing "Stuck" during normal loading).

## How to spot a stuck table

* Card shows a busy border colour (e.g. Ordering red) but the guest count and total show as blank.
* **Stuck** badge appears at the bottom-right of the card.
* Opening the Table Session modal shows: *"No session found for this table."*

<Frame caption="Stuck table card with badge and empty session modal">
  <img src="https://mintlify.s3.us-west-1.amazonaws.com/dishink/images/tables/stuck-tables.png" alt="Stuck table" />
</Frame>

## Wait first, then act

Before force-resetting, wait **10–30 seconds**. Most stuck states self-heal when:

* An offline write completes syncing.
* A different terminal finishes its transaction.
* The tablet re-establishes its Supabase realtime channel.

If the badge is still there after 30 seconds, proceed to reset.

## Force-reset (Manager PIN)

Only managers can force-reset a stuck table:

1. Open the stuck table.
2. In the *"No session found"* body state, tap **Force Reset**.
3. Enter your **Manager PIN**.
4. Confirm.

DishInk clears the table's server-side status back to **Free**. Any orphaned session records are marked resolved in the audit log.

<Warning>
  Force-reset is destructive to unrecovered data. If a session was mid-write and about to sync, force-reset will discard it. Wait first.
</Warning>

## Why stuck tables happen

Common causes:

* **Network drop mid-order.** A waiter added items on an unstable link; the status was queued locally, uploaded, but the session record was retried and dropped.
* **Two terminals conflicting.** Cashier closed a session on Terminal A while Waiter was editing it on Terminal B — the loser's write lands but its session is gone.
* **Server-side glitch during a spike.** Rare, but a partial write during peak load can leave a status without a session.
* **Manual database intervention.** Support or engineering deleting a session for debugging.

## Prevention

* Keep tablets on stable Wi-Fi. See [Network Issues](/troubleshooting/network-issues).
* Avoid closing tables from two terminals simultaneously. Whoever has the modal open should complete or cancel first.
* Report repeat stuck patterns to [Support](/support/contact-support) so we can trace the root cause.

## Tips & best practices

<Tip>
  Establish one rule: **wait 30 seconds, then force-reset.** No exceptions. This eliminates the temptation to reset during an ongoing sync.
</Tip>

<Tip>
  Managers, keep a note if you force-reset. If it happens more than once per shift, escalate to Support.
</Tip>

<Tip>
  Stuck badges on multiple tables at once usually mean a network issue, not a data problem. Check connectivity before resetting.
</Tip>

## Common mistakes

* **Resetting immediately.** The session was probably about to sync. Wait.
* **Non-manager staff attempting reset.** Blocked — they'll be prompted for a Manager PIN.
* **Assuming stuck means "empty and safe."** It might have unsynced items. Waiting preserves them.

## Related pages

* [Table Statuses](/tables/table-statuses)
* [Floor Plan](/tables/floor-plan)
* [Network Issues](/troubleshooting/network-issues)
* [Contact Support](/support/contact-support)