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

# Jobs

> Records of sync and replication work performed by Twine

A **Job** represents a single unit of pipeline work performed by Twine. Every sync run and every replication run is recorded as a Job, and Jobs carry the log output, per-pipeline-step timing metrics, and any error output produced during execution. Each Job is tied to a Domain and optionally to a specific end of a [Domain Mapping](/platform/domain-mappings).

## Job types

* **Sync** - reads entities from a source system, transforms them via property mappings, filters them via conditions, and persists them to Twine. One Sync job is created per sync run.
* **Replication** - sends a single entity to a target system. Filters first, then transforms from Twine's model into the target's model via property mappings, then performs the API calls. One Replication job is created per entity per target.

A Sync job spawns one Replication job per entity, with two exceptions: when Twine itself is the target the flow terminates at persistence, and when a file-based source writes to a file area destination the file is produced as part of the Sync job and no per-entity Replication jobs are spawned.

## Statuses

| Status              | Meaning                                                                                                                                                    |
| ------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `pending`           | Queued and waiting to run.                                                                                                                                 |
| `processing`        | Currently executing.                                                                                                                                       |
| `completed`         | Finished successfully.                                                                                                                                     |
| `skipped`           | Filtered out by conditions before any work was performed.                                                                                                  |
| `failed`            | Errored out. The job log contains the failure detail.                                                                                                      |
| `requeued`          | The target system returned a known error condition; the job was requeued for a later attempt instead of retrying API calls inline.                         |
| `duplicate_aborted` | A duplicate job was detected and this one was cancelled.                                                                                                   |
| `unresolved`        | Exit status was unclear - for example, a hanging job that was manually terminated, or an old uncompleted job picked up by the stale-job garbage collector. |

## Job log

Each Job has a single log that contains everything produced during its run: error output, general log statements, and metrics for how long each step in the pipeline took. The log is the primary surface for debugging job failures.

For instructions on finding and inspecting jobs in the UI, see [Inspecting Jobs](/backoffice/daily-operations/jobs).

<Info>
  This page is a stub.
</Info>