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

# Troubleshooting: page shows nothing — an order or import stuck in an enqueued state

> Diagnose dashboard pages that render an empty list while an order, callback, import, or delta sync says 'enqueued' or 'pending' — the checklist that separates a filter artifact from a queue stuck with no worker picking it up, a backfill gap, and the escalation path when self-serve resets fail.

# Troubleshooting: page shows nothing — an order or import stuck in an enqueued state

Several dashboard surfaces — Inbox → Tickets, Voice → Callbacks, Voice →
Intelligence, Numbers → Connectivity — can show an empty list or a status
stuck on `pending`, `ordered`, or `enqueued`. A page that shows nothing
usually has one of four causes: your filters matched no rows, the item was
enqueued but no worker picked it up, a backfill or delta sync has not run
yet, or the delivery failed downstream. Work through the checks below in
order; each check either solves the page or rules a cause out.

This page is the route-map between those causes. Per-surface remedies are
covered by their own troubleshooting pages — Inbox → Tickets status
transitions are on [Conversation merge](/troubleshooting/conversation-merge),
and Voice → Callbacks scheduling (when the callback confirms but never
fires the call) is on [Callbacks scheduling](/voice/callbacks-scheduling).

## 1. What "enqueued" actually counts as

A record is truly enqueued when it holds a requested-type status —
`requested` on a bulk number-fulfillment job, `pending` on a callback in
queue, `ordered` on a connectivity SIM — and nothing advances it. In the
event log that means a record with the request recorded (or no event at
all), no claim, and no progress heartbeat. A healthy queue keeps that
window short: a `pending` callback becomes `dialing` within seconds, an
`ordered` SIM activates when its scan or eSIM install completes, a
`requested` fulfillment job moves to per-line delivered/failed status.

"Enqueued for a while" reads differently from "enqueued forever." Minutes
on a callback is a queue problem; seconds is normal routing-latency.

## 2. Where empty or stuck surfaces show up — and the self-serve reset

Each surface has its own reset path. Pick the one matching what you see:

* **Inbox → Tickets shows nothing.** The most common cause is a filter
  set narrowing the list to zero — status, assignee, date range, or a
  saved view. Clear the filters. If every filter is cleared and the list
  is still empty while filters say "All," the problem moves to the
  checklist in section 4.
* **Numbers → Connectivity shows a SIM stuck on `ordered`.** For eSIM
  SIMs, activation requires the device-side install to complete — the SIM
  stays `ordered` until the tenant-side finish confirms it. For physical
  SIMs stuck on `ordered` with install done, cancel the order and order
  again. Re-order is safe: only the newest request is live at a time for
  the same ICCID.
* **Voice → Callbacks shows a callback stuck on `pending`.** Cancel the
  entry and create it again with the corrected schedule. If re-created
  callbacks return to `pending` and stay there, the queue itself is not
  moving — go through section 4.
* **Voice → Intelligence shows nothing.** The intelligence data builds
  on a scheduled delta sync; first-time views can render empty until the
  first run completes. If the page stayed empty across full sync cycles,
  go through section 4.

## 3. Queue stuck vs backfill gap vs downstream failure

The three easy-to-confuse causes separate on what the page reports:

* **Queue stuck:** the record exists in a requested-type status
  (`requested` / `pending` / `ordered`) and nothing advances it. The
  import or delivery never started.
* **Backfill gap:** the list is empty, and it starts with "Run a complete
  history scan" or a header that calls the sync as history. Backfill
  imports history — delta sync imports new rows; the two look identical
  from a page that shows neither.
* **Downstream failure:** the record left the requested status and then
  failed — a callback that moved to `dialing` and never connected, an
  import that ran, or an order that completed behind a broken condition.
  A queue reset does not fix this; follow the error the activity reports.

## 4. Order of checks

Run through this before support:

1. **Reset every filter.** Status, assignee, date range, saved views —
   a filter set that matched zero rows shows an empty page indistinguishable
   from a broken one.
2. **Check the role gate.** Some surfaces render scoped to what your role
   can read; a permission-scoped page looks empty because you cannot see,
   not because there is nothing. Ask an admin whether the page is gated
   for you.
3. **Check the time cursor.** Surfaces backed by a scheduled delta sync
   only show data up to the last completed run; recent traffic newer than
   that appears missing until the tick runs.
4. **Page through deliberately.** Pages return in batches (default size
   around 100). On a large dataset, "older than the first page" can read
   as empty.
5. **Check the status page.** If the above found nothing, look at
   `status.orbit.devotel.io` before treating it as your account.

## 5. Escalate when self-serve fails

When a stuck item survives a cancel-then-re-create and the checklist above
rules out filters and the role gate, support can move it from the
platform's dead-letter and replay console — the last-resort recovery path
for an entry no worker advancement picks up. Hand support the stuck
record's ID and the surface it stuck on.

## See also

* [Conversation merge](/troubleshooting/conversation-merge) — Inbox
  status transitions and the related state model.
* [Callbacks scheduling](/voice/callbacks-scheduling) — Voice →
  Callbacks scheduling when the callback confirms but never fires.