Backup and Restore in CockroachDB Cloud

CockroachDB Cloud provides comprehensive backup and restore capabilities to protect your data and enable disaster recovery.

Overview

CockroachDB Cloud offers two types of backups:

Managed Backups: Automated backups managed by Cockroach Labs
Self-Managed Backups: Manual backups you control and store in your own cloud storage

Managed Backups

All CockroachDB Cloud clusters include automatic managed backups at no additional cost (storage costs apply for some plans).

Managed Backup Features

Automatic scheduling: Backups run automatically based on configured frequency
Retention management: Old backups are automatically removed based on retention policy
Secure storage: Backups stored in Cockroach Labs-managed cloud storage
Point-in-time recovery: Restore to specific timestamps (with revision history)
Incremental backups: Efficient storage using incremental backup chains

Backup Settings by Plan

Basic
Standard
Advanced

Basic Cluster Backups:

Frequency: Daily (not configurable)
Retention: 30 days (not configurable)
Storage: Included in RU pricing
Restoration: To same cluster or new cluster

Configure Managed Backups

Navigate to Backups

Select your cluster
Click Backup and Restore in the left navigation

Adjust Settings

Click Settings and configure:Frequency:

Daily (default)
Every 4 hours
Every 1 hour

Retention:

Enter number of days (1-365)
Default: 30 days

Save Configuration

Click Save to apply changes

Backup frequency and retention changes apply to new backups. Existing backups follow their original retention schedule.

View Managed Backups

Access backup history and details:

Go to Backup List

Navigate to Backup and Restore page for your cluster

Review Backups

The list shows:

Backup timestamp
Backup type (full or incremental)
Size
Status
Expiration date

View Details

Click any backup to see:

Backup UUID
Databases included
Backup chain information
Download connection details (for recovery)

Restore from Managed Backup

Restore your cluster from a managed backup:

Select Backup

Go to Backup and Restore page
Find the backup to restore
Click Restore

Choose Restore Target

Select where to restore:

This cluster: Overwrites current data
New cluster: Creates new cluster with backup data

Configure Restore

For new cluster restoration:

Name the new cluster
Select plan and configuration
Review cost estimate

Start Restoration

Review restore details
Click Restore
Monitor progress on the Backups page

Restoring to existing cluster: This operation will overwrite all current data. Ensure you have a recent backup before proceeding.

Point-in-Time Recovery

Restore to a specific timestamp (requires revision history):

Initiate Restore

Click Restore on a backup with revision history

Select Timestamp

Choose Point-in-time
Enter desired timestamp or use date picker
Must be within backup’s revision history window

Complete Restore

Select target and click Restore

Self-Managed Backups

Take manual backups to your own cloud storage for additional control.

Benefits of Self-Managed Backups

Full control: You manage backup storage and lifecycle
Custom schedules: Create backup schedules via SQL
Cloud storage: Use your AWS S3, GCS, or Azure Blob storage
Encryption: Optional encryption with KMS or passphrase
Locality-aware: Partition backups by node locality

Set Up Cloud Storage

Configure authentication to your cloud storage:

AWS S3
Google Cloud Storage
Azure Blob Storage

Using IAM Role (Recommended):

BACKUP DATABASE bank 
INTO 's3://bucket-name/backup-path?AUTH=implicit&AWS_ACCESS_KEY_ID={key}&AWS_SECRET_ACCESS_KEY={secret}'

Using Access Keys:

BACKUP DATABASE bank 
INTO 's3://bucket-name/backup-path?AWS_ACCESS_KEY_ID={key}&AWS_SECRET_ACCESS_KEY={secret}'

Using Service Account:

BACKUP DATABASE bank 
INTO 'gs://bucket-name/backup-path?AUTH=implicit'

With Credentials:

BACKUP DATABASE bank 
INTO 'gs://bucket-name/backup-path?CREDENTIALS={base64-encoded-key}'

BACKUP DATABASE bank 
INTO 'azure://bucket-name/backup-path?AZURE_ACCOUNT_NAME={account}&AZURE_ACCOUNT_KEY={key}'

Create Full Backup

Create a complete backup of your data:

-- Backup entire cluster
BACKUP INTO 's3://bucket/full-backup-path?AUTH=implicit';

-- Backup specific database
BACKUP DATABASE mydb INTO 's3://bucket/db-backup-path?AUTH=implicit';

-- Backup specific tables
BACKUP TABLE mydb.table1, mydb.table2 
INTO 's3://bucket/table-backup-path?AUTH=implicit';

Create Incremental Backup

Incremental backups save only changes since the last backup:

-- Create incremental backup
BACKUP INTO LATEST IN 's3://bucket/backup-path?AUTH=implicit';

Schedule Automated Backups

Create a backup schedule:

-- Daily full backups at 2 AM UTC, hourly incrementals
CREATE SCHEDULE daily_backup
  FOR BACKUP INTO 's3://bucket/scheduled-backup?AUTH=implicit'
  RECURRING '@daily'
  FULL BACKUP '@weekly'
  WITH SCHEDULE OPTIONS ignore_existing_backups;

Schedule Options:

@hourly, @daily, @weekly, @monthly
Cron expressions: 0 2 * * * (2 AM daily)
FULL BACKUP cadence for full vs incremental

Backup with Revision History

Enable point-in-time recovery:

BACKUP DATABASE bank 
INTO 's3://bucket/backup-path?AUTH=implicit'
WITH revision_history;

Revision history allows restoration to any point within the garbage collection window.

Encrypted Backups

Encrypt backups for additional security:

Passphrase
AWS KMS
GCP KMS

BACKUP DATABASE bank 
INTO 's3://bucket/backup?AUTH=implicit'
WITH encryption_passphrase = 'my-secret-passphrase';

BACKUP DATABASE bank 
INTO 's3://bucket/backup?AUTH=implicit&AWS_ACCESS_KEY_ID={key}&AWS_SECRET_ACCESS_KEY={secret}'
WITH kms = 'aws:///arn:aws:kms:region:account:key/key-id';

BACKUP DATABASE bank 
INTO 'gs://bucket/backup?AUTH=implicit'
WITH kms = 'gs:///projects/project/locations/region/keyRings/ring/cryptoKeys/key';

Restore from Self-Managed Backup

Restore data from your cloud storage:

-- Restore full database
RESTORE DATABASE bank 
FROM LATEST IN 's3://bucket/backup-path?AUTH=implicit';

-- Restore specific tables
RESTORE TABLE bank.accounts 
FROM LATEST IN 's3://bucket/backup-path?AUTH=implicit';

-- Restore to specific timestamp
RESTORE DATABASE bank 
FROM LATEST IN 's3://bucket/backup-path?AUTH=implicit'
AS OF SYSTEM TIME '2024-01-15 10:00:00';

Restores run as jobs. Monitor progress with SHOW JOBS.

Backup Monitoring

Monitor backup jobs and health:

View Backup Jobs

-- Show all backup jobs
SHOW JOBS SELECT * FROM [SHOW JOBS] WHERE job_type = 'BACKUP';

-- Show running backup jobs
SHOW JOBS SELECT * FROM [SHOW JOBS] 
WHERE job_type = 'BACKUP' AND status = 'running';

View Backup Details

-- Show backup contents
SHOW BACKUP LATEST IN 's3://bucket/backup-path?AUTH=implicit';

-- Show backup with details
SHOW BACKUP LATEST IN 's3://bucket/backup-path?AUTH=implicit'
WITH check_files;

Monitor in Cloud Console

Track backups in the Cloud Console:

Backup and Restore page: View managed backup history
Jobs page: Monitor backup job progress
Metrics page: Backup-related metrics
Alerts: Configure alerts for backup failures

Backup Best Practices

Retention Policy

Recommended Retention

Production: 30-90 days
Compliance: Based on regulatory requirements
Development: 7-14 days
Cost vs. Risk: Balance storage costs with recovery needs

Backup Frequency

Frequency Recommendations:

Critical data: Hourly backups
Production: Daily backups
Development: Daily or less frequent
Consider: RPO (Recovery Point Objective) requirements

Testing Restores

Regularly test backup restoration:

Schedule Tests

Test restores monthly or quarterly

Validate Data

Verify restored data integrity and completeness

Measure Time

Track restoration time for disaster recovery planning

Document Process

Maintain runbooks for restore procedures

Multiple Backup Strategies

Implement defense in depth:

Use both managed and self-managed backups
Store backups in multiple regions
Keep backups in different cloud providers
Maintain offline backup copies for critical data

Disaster Recovery

Recovery Point Objective (RPO)

RPO is the maximum acceptable data loss:

Backup Frequency	RPO
Hourly	1 hour
Daily	24 hours
Weekly	7 days

Recovery Time Objective (RTO)

RTO is the maximum acceptable downtime: Factors Affecting RTO:

Backup size
Network bandwidth
Cluster capacity
Restore complexity

Typical Restore Times:

Small clusters (under 100 GB): 15-30 minutes
Medium clusters (100GB-1TB): 1-3 hours
Large clusters (over 1TB): Several hours

DR Planning

Define Requirements

Establish RPO and RTO requirements

Design Strategy

Choose backup frequency and retention

Test Regularly

Perform restore tests and measure times

Document Procedures

Create detailed recovery runbooks

Review and Update

Revisit DR plan quarterly

Troubleshooting

Backup Failed

Common Causes:

Insufficient storage space
Authentication failures
Network connectivity issues
Concurrent schema changes

Solutions:

Check error message in job details
Verify cloud storage credentials
Ensure adequate storage space
Retry backup after resolving issues

Restore Failed

Common Causes:

Incompatible CockroachDB versions
Corrupt backup files
Insufficient cluster capacity
Schema conflicts

Solutions:

Verify backup integrity with SHOW BACKUP
Check cluster has sufficient capacity
Review CockroachDB version compatibility
Contact Support if issues persist

Next Steps

Monitor Backups

Set up backup monitoring and alerts

Cloud Storage

Configure cloud storage authentication

Backup SQL

Learn detailed backup SQL syntax

Disaster Recovery

Plan comprehensive DR strategy

Getting Started

Cluster Management

Security

Billing & Plans

​Overview

​Managed Backups

​Managed Backup Features

​Backup Settings by Plan

​Configure Managed Backups

​View Managed Backups

​Restore from Managed Backup

​Point-in-Time Recovery

​Self-Managed Backups

​Benefits of Self-Managed Backups

​Set Up Cloud Storage

​Create Full Backup

​Create Incremental Backup

​Schedule Automated Backups

​Backup with Revision History

​Encrypted Backups

​Restore from Self-Managed Backup

​Backup Monitoring

​View Backup Jobs

​View Backup Details

​Monitor in Cloud Console

​Backup Best Practices

​Retention Policy

Recommended Retention

​Backup Frequency

​Testing Restores

​Multiple Backup Strategies

​Disaster Recovery

​Recovery Point Objective (RPO)

​Recovery Time Objective (RTO)

​DR Planning

​Troubleshooting

​Backup Failed

​Restore Failed

​Next Steps

Monitor Backups

Cloud Storage

Backup SQL

Disaster Recovery

Build docs developers (and LLMs) love

Overview

Managed Backups

Managed Backup Features

Backup Settings by Plan

Configure Managed Backups

View Managed Backups

Restore from Managed Backup

Point-in-Time Recovery

Self-Managed Backups

Benefits of Self-Managed Backups

Set Up Cloud Storage

Create Full Backup

Create Incremental Backup

Schedule Automated Backups

Backup with Revision History

Encrypted Backups

Restore from Self-Managed Backup

Backup Monitoring

View Backup Jobs

View Backup Details

Monitor in Cloud Console

Backup Best Practices

Retention Policy

Backup Frequency

Testing Restores

Multiple Backup Strategies

Disaster Recovery

Recovery Point Objective (RPO)

Recovery Time Objective (RTO)

DR Planning

Troubleshooting

Backup Failed

Restore Failed

Next Steps