> For the complete documentation index, see [llms.txt](https://docs.silk.us/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.silk.us/manage/protect-your-environment/replication-and-recovery.md).

# Replication and Recovery

Learn how Silk creates recovery points, replicates volume groups between DataPods, and supports local recovery and disaster recovery failover.

Silk provides storage-level snapshots and replication to support data recovery.

{% columns %}
{% column %}
**Snapshots** preserve point-in-time copies of a volume group on a DataPod.
{% endcolumn %}

{% column %}
**Replication** transfers snapshot changes to another DataPod, providing recovery points at a separate location.
{% endcolumn %}
{% endcolumns %}

The right recovery path depends on whether you need to restore data locally or run an application using replicated volumes on the target DataPod.&#x20;

{% hint style="info" %}
**Database-level replication**

You can also use your database's native replication capabilities with Silk for synchronous or asynchronous replication, where supported by the database. Database-level replication is configured and managed through the database platform - the Silk replication described in this article operates at the volume group level between DataPods. Choose the approach that fits your application's consistency and recovery requirements.&#x20;
{% endhint %}

## Choose a recovery path

| Need                                                               | Silk capability                                   | What it does                                                                                             |
| ------------------------------------------------------------------ | ------------------------------------------------- | -------------------------------------------------------------------------------------------------------- |
| Inspect an earlier state without changing the original volumes     | Create a view from a snapshot                     | Provides a writable copy that can be mapped to a host for validation or recovery work                    |
| Return data to an earlier point on the same DataPod                | Restore a snapshot                                | Writes the selected snapshot's data back to its base volumes, overwriting their current data.            |
| Continue operations when the source DataPod or site is unavailable | Fail over a disaster recovery replication session | Restores a replicated snapshot on the target DataPod and make its volumes available for application use. |

Snapshots are created at the **volume group** level, so a snapshot captures the volumes in that group together. Snapshot schedules can create regular recovery points. For applications spread across multiple volume groups, Silk also supports cross-volume-group snapshots to capture coordinated points in time.

A snapshot isn't directly mapped to a host. Create a **view** when you need to access its data without restoring it over the original volumes.&#x20;

## Replicate volume groups between DataPods

Silk replication is asynchronous and snapshot based. A replication session connects a source volume group to a corresponding volume group on a peer DataPod. After the initial synchronization, Silk transfers changes captured in subsequent replication snapshots rather than copying the entire volume group each cycle. All volumes in the source volume group participate in that session.

A **replication peer** establishes the connection between DataPods, while a **replication session** specifies the volume group, replication type, desired recovery point objective (RPO), target exposure, and snapshot retention. The DataPods need working management and replication network paths before a session can begin.&#x20;

Silk supports these session types:

{% tabs %}
{% tab title="Disaster recovery (DR)" %}
Maintains a source and target relationship for failover.&#x20;

A volume group can have one DR replication session.&#x20;
{% endtab %}

{% tab title="One-way replication" %}
Sends changes from source to target without a DR pairing. It can also support cascading replication, in which a target replicates onward to another DataPod.&#x20;
{% endtab %}
{% endtabs %}

The **desired RPO** expresses the maximum acceptable gap between the latest replicated recovery point and the source's current data. A shorter target calls for more frequent replication and may use more network bandwidth. A configured RPO is a target - check the session status and RPO history in the SDP Management Console to confirm that replication is meeting it.&#x20;

## Plan what happens at the target

When configuring a session, choose how target volumes are exposed:

<table data-view="cards"><thead><tr><th align="center"></th><th align="center"></th><th></th></tr></thead><tbody><tr><td align="center"><h2><i class="fa-eye-slash">:eye-slash:</i></h2></td><td align="center"><strong>Mapped - Not Exposed</strong></td><td>Target volumes cannot be accessed by hosts during replication.</td></tr><tr><td align="center"><h2><i class="fa-eye">:eye:</i></h2></td><td align="center"><strong>Read Only</strong></td><td>Hosts can access target volumes without writing to them. </td></tr></tbody></table>

{% hint style="info" %}
Read-only exposure can reduce setup work during recovery, but volumes configured this way remain read only after failover until an administrator changes their permissions.
{% endhint %}

Also plan snapshot retention and available capacity on both DataPods. Snapshots and their dependent views consume capacity as data changes, and insufficient quota can suspend a replication session. Confirm that the target environment has the host connectivity and application configuration needed to use its volumes during recovery.&#x20;

## Recover from a replicated snapshot

Silk doesn't automatically decide that an application or source DataPod has failed. An administrator or external automation initiates **DR failover on the target DataPod** and selects an available replicated snapshot, or the latest available recovery point.

During failover, the target suspends incoming replication updates, restores the selected snapshot to its target volumes, and exposes teh restored data to hosts. The application must then be brought up against those volumes using the organization's recovery plan. Data written to the source after the selected recovery point isn't part of the restored image.&#x20;


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.silk.us/manage/protect-your-environment/replication-and-recovery.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
