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

# Disaster Recovery Overview

export const InternalLink = ({version, path = "", children, ...props}) => {
  let detectedVersion = version || "stable";
  if (typeof window !== 'undefined' && !version) {
    const match = window.location.pathname.match(/\/docs\/([^/]+)/);
    if (match) {
      detectedVersion = match[1];
    }
  }
  const normalizedPath = path.startsWith("/") ? path.slice(1) : path;
  return <a href={`/docs/${detectedVersion}/${normalizedPath}`} {...props}>
      {children}
    </a>;
};

Resilient deployments aim for continuity in database operations to protect from data loss and downtime. To maintain resiliency, it is necessary to build deployments with *high availability* and *disaster recovery* coverage.

* [High availability](#choose-a-high-availability-strategy): Continuously access data without interruption even in the presence of failures or disruptions to maximize uptime.
* [Disaster recovery](#choose-a-disaster-recovery-strategy): Recover from a major incident or disaster to minimize downtime and data loss.

To build resilient cluster deployments, CockroachDB provides both <InternalLink path="architecture/replication-layer">built-in replication</InternalLink> for high availability and disaster recovery features for coverage during unplanned incidents.

<Tip>
  For a practical guide on how CockroachDB replicates, distributes, and rebalances data, refer to the <InternalLink path="demo-replication-and-rebalancing">Replication and Rebalancing</InternalLink> demo.
</Tip>

## Resilience strategy

As you evaluate CockroachDB's disaster recovery features, consider your organization's requirements for the amount of tolerable data loss and the acceptable length of time to recover.

* Recovery Point Objective (RPO): The maximum amount of data loss (measured by time) that an organization can tolerate.
* Recovery Time Objective (RTO): The maximum length of time it should take to restore normal operations following an outage.

For example, when you use backups:

<img src="https://mintcdn.com/cockroachlabs/Z2623UEgCBYdavu9/images/v24.1/rpo-rto.png?fit=max&auto=format&n=Z2623UEgCBYdavu9&q=85&s=c668fd60c9d7bdecfaa507582a530e31" alt="Simulating RPO and RTO. With RPO representing the tolerable data loss, and RTO representing the tolerable time to recovery." width="918" height="320" data-path="images/v24.1/rpo-rto.png" />

For a comparison of CockroachDB resiliency features, refer to the following sections:

* [Choose a high availability strategy](#choose-a-high-availability-strategy)
* [Choose a disaster recovery strategy](#choose-a-disaster-recovery-strategy)

### Choose a high availability strategy

CockroachDB uses synchronous, built-in replication to create and distribute copies of data, ensuring consistency across the data copies. The database can tolerate nodes going offline without service interruption, whether in a single-region or <InternalLink path="multiregion-overview">multi-region</InternalLink> cluster.

|                     | Single-region replication (synchronous)                                                     | Multi-region replication (synchronous)                                                                       |
| ------------------- | ------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------ |
| **RPO**             | 0 seconds                                                                                   | 0 seconds                                                                                                    |
| **RTO**             | Zero RTO Potential increased latency for 1-9 seconds                                        | Zero RTO Potential increased latency for 1-9 seconds                                                         |
| **Write latency**   | Region-local write latency p50 latency \< 5ms (for multiple availability zones in us-east1) | Cross-region write latency p50 latency > 50ms (for a multi-region cluster in us-east1, us-east-2, us-west-1) |
| **Recovery**        | Automatic                                                                                   | Automatic                                                                                                    |
| **Hardware cost**   | 1x Networking between availability zones                                                    | 1x Networking between datacenters                                                                            |
| **Minimum regions** | 1                                                                                           | 3                                                                                                            |
| **Fault tolerance** | Zero RPO node, availability zone failures                                                   | Zero RPO node, availability zone failures, region failures                                                   |

For details on designing your cluster topology for high availability with replication, refer to the <InternalLink path="disaster-recovery-planning#hardware-failure">Disaster Recovery Planning</InternalLink> page.

### Choose a disaster recovery strategy

CockroachDB is designed to recover automatically; however, building <InternalLink path="backup-and-restore-overview">backups</InternalLink> or <InternalLink path="physical-cluster-replication-overview">physical cluster replication</InternalLink> into your disaster recovery planning is an important part of a resilient deployment.

|                     | Point in time backup & restore                               | Physical cluster replication (asynchronous)                          |
| ------------------- | ------------------------------------------------------------ | -------------------------------------------------------------------- |
| **RPO**             | >=5 minutes                                                  | 10s of seconds                                                       |
| **RTO**             | Minutes to hours, depending on data size and number of nodes | Seconds to minutes, depending on cluster size                        |
| **Write latency**   | No impact                                                    | No impact                                                            |
| **Recovery**        | Manual restore                                               | Manual failover                                                      |
| **Hardware cost**   | 1x + S3-like storage & networking                            | 2x + networking between datacenters                                  |
| **Minimum regions** | 1                                                            | 2                                                                    |
| **Fault tolerance** | Not applicable                                               | Zero RPO node, availability zone, region failure with loss up to RPO |