> For the complete documentation index, see [llms.txt](https://docs.vega365.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.vega365.com/tandem/configuration/on-the-management-client/backfill-configuration.md).

# Backfill Configuration

Enabling backfill on each recording server, choosing its staging drive, and throttling its bandwidth.

## Backfill configuration

Configure backfill settings per recording server in the Management Client: MIP Plug-ins → Tandem → BackFill. The node has two tabs: **Flow Dashboard** (monitoring) and **Configuration** (settings).

**Backfill is bidirectional.** Either member of a cluster can fall behind, and either can recover from the other. The primary backfills from the secondary, and the secondary backfills from the primary, depending on which one was down.&#x20;

{% hint style="info" %}
**A cluster must exist first.** Until recorders are clustered, the BackFill node reports that no Tandem clusters are configured and directs you to **Tandem → Clustering → Configuration**. BackFill becomes available once you create the cluster. See Clustering.
{% endhint %}

{% hint style="warning" %}
**Backfill requires unencrypted media.** Media encrypted on one recording server cannot be decrypted on another, so backfill cannot copy recordings between cluster members when XProtect's media encryption is enabled. When encryption at rest is required, Tandem runs without backfill—both recorders still record independently, and failover is unchanged, but content missed during an outage cannot be recovered. See [System Requirements](https://docs.vega365.com/tandem/system-requirements.md).
{% endhint %}

<figure><img src="https://92057524-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi0mCLda4cGtI3Jwb6IQA%2Fuploads%2FnWri940XYB8X0OjJvxce%2Fconfig_cluster_first.png?alt=media&amp;token=c6a3e7f8-5e5a-4b1e-8f24-7974945ef9da" alt=""><figcaption><p>The Backfill Flow Dashboard reporting that no Tandem clusters are configured</p></figcaption></figure>

### Flow Dashboard

The Flow Dashboard shows the clustered recorders as primary and secondary servers, with the data flow between them. Use this dashboard to visualize server status and data flow.

<figure><img src="https://92057524-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi0mCLda4cGtI3Jwb6IQA%2Fuploads%2FsZSBfe2ujT26lMFF8Bpa%2FAfter_clustering.png?alt=media&amp;token=0e032cb2-3234-40df-860f-b5534a07ed55" alt=""><figcaption><p>The Server Flow Dashboard showing a primary and secondary recorder</p></figcaption></figure>

Counters at the top summarize how many servers are online, syncing, and offline. Each server card shows the state of its BackFill service, its camera count, and whether backfill is enabled. Use the search box to find a recorder by name.

| State        | Meaning                                                      |
| ------------ | ------------------------------------------------------------ |
| Online       | The recorder is reachable and responsive to backfill control |
| Syncing      | The recorder is fixing missing content                       |
| Offline      | The recording service is not reachable                       |
| Service Down | The BackFill service is not reachable                        |
| Fetching     | Communication is in progress — wait for it to finish.        |

### Configuration tab

The Configuration tab lists the recording servers and shows how many are online. Filter by **All**, **Running**, **Syncing**, **Stopped,** or **Restarting**, or search by name. **Refresh Servers** re-reads the list.

<figure><img src="https://92057524-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi0mCLda4cGtI3Jwb6IQA%2Fuploads%2F8OqrnIWo3MmL3Ck7zci4%2Fconfig_dashboard.png?alt=media&amp;token=f776ac16-8def-48f7-a7fe-5eb63fde11ff" alt=""><figcaption><p>The BackFill Configuration tab listing two recording servers</p></figcaption></figure>

Each recording server shows its name, its role in the cluster, and its service state, with these figures:

| Figure    | Meaning                                                         |
| --------- | --------------------------------------------------------------- |
| Cameras   | Cameras on that recorder                                        |
| Service   | State of the BackFill service                                   |
| Uptime    | How long the service has been running                           |
| Backfills | Number of backfill jobs the recorder has run, over its lifetime |
| Data      | Volume backfilled over its lifetime                             |

Backfills and Data are lifetime totals, not figures for the current outage or the current day.

#### Per-recorder controls

Each recorder carries **Restart** and **Refresh** buttons, a **Settings** button, and a **…** menu:

<figure><img src="https://92057524-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi0mCLda4cGtI3Jwb6IQA%2Fuploads%2FgxkSHTTbzJm3bXlo3DE0%2FBackfill_service_controls_per_recorder.png?alt=media&amp;token=31ef0650-4b06-4b0c-b5ff-b6bdb1c8e6f5" alt=""><figcaption></figcaption></figure>

#### The "..." menu

| Control           | What it does                                               |
| ----------------- | ---------------------------------------------------------- |
| Expand / Collapse | Shows or hides the recorder's detail                       |
| Restart Service   | Restarts the BackFill service on that recorder             |
| Refresh Status    | Re-reads the recorder's current state                      |
| Copy Host / IP    | Copies the recorder's hostname or address to the clipboard |
| Rollback Backfill | Removes a backfill job's content from the recorder         |

{% hint style="danger" %}
**Rollback Backfill deletes recorded video.** Select a backfill job to remove its content from the recorder, including content already merged into the timeline. The video it restored is then gone from that recorder. Use it only when you intend to discard what a backfill job brought in.
{% endhint %}

### Backfill settings

Select the settings button on a recording server to open its settings.

<figure><img src="https://92057524-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi0mCLda4cGtI3Jwb6IQA%2Fuploads%2FX2e6odZ0yPfitp612csa%2FPer_Recorder_Config_and_Scheduling.png?alt=media&amp;token=69f9981f-6677-45a1-aecc-b51b8dce1d48" alt=""><figcaption><p>Per-recorder backfill settings</p></figcaption></figure>

Settings apply to the recording server where you open them. Configure both members of a cluster—because backfill runs in either direction, a member left unconfigured cannot recover its own gaps.

Select **Save Settings** to apply, or **Cancel** to discard.

#### Enable Backfill

Turns backfill on or off for that recording server. With it off, the server does not recover missed content after an outage, and gaps in its timeline remain.

#### Port

The port the BackFill service uses. It must match on both members of the cluster, and must be open between them in the firewall. The Tandem software in the Management client also uses this port to communicate with each recorder. Changing it on one recorder without changing it on its peer — and without updating the firewall rule — stops backfill between them. See [Network Requirements](https://docs.vega365.com/tandem/network-requirements.md).

#### Drive

The drive where backfilled content is **staged** before it is merged into the recorder's timeline. Every drive on the recording server is listed; you can't set a specific folder path.

The drive must have spare capacity for the staged content. This is what determines how long an outage the recorder can recover from.

Worked example. Suppose the staging drive is the same drive letter as the archive drive, and the recorder was down for one day. That drive must have a day's spare capacity for **all** cameras across all storage profiles—live and archive—that write to that drive letter. If it does not, backfill fails for that duration.

{% hint style="warning" %}
**Spare capacity sets your recoverable outage window.** A drive with two days of headroom can recover a two-day outage and no more. Size the staging drive against the longest outage you expect to recover from. Remember, the recorder's own pre-outage recordings are still on it, so the space doesn't free up while it is down.&#x20;
{% endhint %}

Changes to the drive take effect at the next backfill. No service restart is required.

#### Bandwidth throttling

The throttling table limits the rate at which the **recording server provides data** to its peer. Each row covers a range of hours and a limit in MB/s (megabytes per second).

| Field        | Meaning                         |
| ------------ | ------------------------------- |
| Start Hour   | First hour the limit applies    |
| End Hour     | Hour the limit runs to          |
| Limit (MB/s) | Maximum rate during those hours |

Use **Add** and **Delete** to manage rows. The rules:

* **Rows cannot overlap.** Each hour is covered by at most one row.
* **Hours with no row are unlimited.** The recorder uses all available bandwidth during any hour the table does not cover.
* **The table can be empty.** An empty table means no throttling at any time.
* **Changes take effect at the next backfill job.** A job already running continues at the rate it started with.

Throttling trades recovery speed for bandwidth. A tight limit during working hours keeps backfill away from camera ingest and client traffic; a wide or absent limit overnight lets the recorder catch up faster.
