Backup and Restore

The Redis Operator provides two CRDs for backup management: RedisBackup for on-demand backups and RedisScheduledBackup for automated backup schedules.

Backup Methods

Two backup methods are supported:

RDB (Redis Database)

Point-in-time snapshot of the entire dataset. Recommended for most use cases.

Format: Binary RDB file
Size: Compact, compressed
Speed: Fast for restore
Consistency: Atomic snapshot

AOF (Append-Only File)

Log of all write operations. Useful for audit trails or granular recovery.

Format: Text-based command log archive
Size: Larger than RDB
Speed: Slower restore (replays commands)
Consistency: Per-operation durability

On-Demand Backups

Create a RedisBackup resource to trigger an immediate backup.

Basic Backup

backup.yaml

apiVersion: redis.io/v1
kind: RedisBackup
metadata:
  name: my-cluster-backup-001
  namespace: default
spec:
  clusterName: my-cluster
  target: prefer-replica
  method: rdb
  destination:
    s3:
      bucket: redis-backups
      path: production/my-cluster
      region: us-east-1

Create the backup

kubectl apply -f backup.yaml

Monitor progress

kubectl get redisbackup my-cluster-backup-001 -o wide

Expected phases: Pending → Running → Completed

Check backup details

kubectl get redisbackup my-cluster-backup-001 -o yaml

Inspect status.backupPath, status.backupSize, and status.completedAt.

Backup Targets

The target field controls which pod executes the backup:

prefer-replica (default) - Prefer a replica to avoid impact on primary
primary - Always run on the primary pod

Backups always run on a single pod. The operator selects an available replica if prefer-replica is set, falling back to primary if no replicas are available.

S3-Compatible Storage

Backups are stored in S3-compatible object storage. Configure credentials via a Secret:

backup-credentials.yaml

apiVersion: v1
kind: Secret
metadata:
  name: backup-credentials
  namespace: default
type: Opaque
stringData:
  AWS_ACCESS_KEY_ID: AKIAIOSFODNN7EXAMPLE
  AWS_SECRET_ACCESS_KEY: wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY

Reference the secret in your RedisCluster spec:

cluster.yaml

apiVersion: redis.io/v1
kind: RedisCluster
metadata:
  name: my-cluster
spec:
  instances: 3
  backupCredentialsSecret:
    name: backup-credentials
  # ... other fields

Never commit backup credentials to version control. Use sealed secrets, external secrets operator, or cloud-native secret management.

Scheduled Backups

Automate backups with RedisScheduledBackup using cron expressions.

Daily Backup Schedule

scheduled-backup.yaml

apiVersion: redis.io/v1
kind: RedisScheduledBackup
metadata:
  name: my-cluster-daily
  namespace: default
spec:
  clusterName: my-cluster
  schedule: "0 2 * * *"  # 2 AM daily
  target: prefer-replica
  method: rdb
  destination:
    s3:
      bucket: redis-backups
      path: production/my-cluster
      region: us-east-1
  successfulBackupsHistoryLimit: 7
  failedBackupsHistoryLimit: 3

Schedule Syntax

Use standard cron format:

┌───────────── minute (0 - 59)
│ ┌───────────── hour (0 - 23)
│ │ ┌───────────── day of month (1 - 31)
│ │ │ ┌───────────── month (1 - 12)
│ │ │ │ ┌───────────── day of week (0 - 6) (Sunday to Saturday)
│ │ │ │ │
* * * * *

Examples:

0 2 * * * - Daily at 2 AM
0 */6 * * * - Every 6 hours
0 0 * * 0 - Weekly on Sunday at midnight
0 0 1 * * - Monthly on the 1st at midnight

History Limits

Control backup retention with history limits:

successfulBackupsHistoryLimit: 7   # Keep last 7 successful backups
failedBackupsHistoryLimit: 3       # Keep last 3 failed backups

Older backups are automatically deleted by the operator. Object storage artifacts are not automatically cleaned up.

The operator only deletes RedisBackup CRs, not the backup files in object storage. Implement lifecycle policies in your S3 bucket for automatic cleanup.

Suspend Scheduled Backups

Temporarily pause scheduled backups:

kubectl patch redisscheduledbackup my-cluster-daily \
  --type merge \
  -p '{"spec":{"suspend":true}}'

Resume:

kubectl patch redisscheduledbackup my-cluster-daily \
  --type merge \
  -p '{"spec":{"suspend":false}}'

Monitoring Backups

List All Backups

kubectl get redisbackups -n default

Check Backup Status

kubectl describe redisbackup my-cluster-backup-001

Look for events indicating backup start, progress, and completion.

Metrics

The operator exposes backup metrics:

redis_last_successful_backup_timestamp{cluster="my-cluster"} - Timestamp of last successful backup
redis_backup_phase_count{phase="Completed"} - Count of completed backups
redis_backup_phase_count{phase="Failed"} - Count of failed backups

Alerts

The default PrometheusRule includes a RedisBackupMissing alert that fires when no successful backup has completed within 24 hours (configurable).

Restore from Backup

Restore a cluster from a backup using the bootstrap field.

Restore only works during cluster creation. You cannot restore into an existing cluster with data.

Create Cluster from Backup

restore-cluster.yaml

apiVersion: redis.io/v1
kind: RedisCluster
metadata:
  name: my-cluster-restored
  namespace: default
spec:
  instances: 3
  imageName: redis:7.2
  storage:
    size: 10Gi
  backupCredentialsSecret:
    name: backup-credentials
  bootstrap:
    backupName: my-cluster-backup-001

Verify backup exists

kubectl get redisbackup my-cluster-backup-001 -o yaml

Confirm status.phase is Completed and status.backupPath is set.

Create the cluster

kubectl apply -f restore-cluster.yaml

Monitor restore progress

kubectl get rediscluster my-cluster-restored -o wide -w

Wait for phase to reach Healthy.

Verify data

Connect to the restored cluster and verify data integrity:

kubectl exec -it my-cluster-restored-0 -- redis-cli
> DBSIZE
> GET sample-key

Backup Workflow Details

Understanding the internal backup process:

Cluster validation

Operator verifies the target cluster exists and is in Healthy phase.

Pod selection

Based on target field, operator selects a replica (if available) or primary pod.

Backup trigger

Operator sends POST /v1/backup to the instance manager on the target pod at internal/controller/backup/executor.go:26.

RDB/AOF generation

Instance manager executes:

RDB: Triggers BGSAVE or copies latest dump.rdb
AOF: Packages AOF files into an archive

Upload to S3

Instance manager uploads the backup artifact to the configured S3 destination.

Status update

Operator updates RedisBackup status with:

backupPath - S3 key/path
backupSize - Size in bytes
artifactType - rdb or aof-archive
completedAt - Completion timestamp

Best Practices

Backup Timing

Schedule backups during low-traffic periods (e.g., 2-4 AM)
Use prefer-replica to avoid primary impact
Avoid overlapping backup windows

Retention Strategy

Production: 7+ daily backups, 4+ weekly backups, 12 monthly backups
Staging: 3-7 daily backups
Development: 1-3 daily backups

Disaster Recovery

Store backups in a different region than the cluster
Test restore procedures quarterly
Document RPO (Recovery Point Objective) and RTO (Recovery Time Objective)
Keep at least one off-cluster backup for catastrophic failures

S3 Bucket Configuration

lifecycle-policy.json

{
  "Rules": [
    {
      "Id": "ExpireOldBackups",
      "Status": "Enabled",
      "Expiration": {
        "Days": 30
      },
      "Filter": {
        "Prefix": "production/"
      }
    },
    {
      "Id": "TransitionToGlacier",
      "Status": "Enabled",
      "Transitions": [
        {
          "Days": 7,
          "StorageClass": "GLACIER"
        }
      ]
    }
  ]
}

Enable S3 versioning and cross-region replication for production backups.

Troubleshooting

Backup Stuck in Pending

Cause: Cluster not in Healthy phase. Solution: Check cluster status and resolve any issues:

kubectl get rediscluster my-cluster -o jsonpath='{.status.phase}'
kubectl describe rediscluster my-cluster

Backup Failed with S3 Error

Cause: Invalid credentials or bucket permissions. Solution: Verify credentials and IAM policy:

kubectl get secret backup-credentials -o yaml

Required IAM permissions:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "s3:PutObject",
        "s3:GetObject",
        "s3:ListBucket"
      ],
      "Resource": [
        "arn:aws:s3:::redis-backups/*",
        "arn:aws:s3:::redis-backups"
      ]
    }
  ]
}

Restore Not Working

Cause: Restore can only run on a new cluster, not existing clusters. Solution: Delete the existing cluster first or create a new cluster with a different name:

kubectl delete rediscluster my-cluster
kubectl apply -f restore-cluster.yaml

Get Started

Core Concepts

Configuration

Operations

Runbooks

Backup Methods

RDB (Redis Database)

AOF (Append-Only File)

On-Demand Backups

Basic Backup

Backup Targets

S3-Compatible Storage

Scheduled Backups

Daily Backup Schedule

Schedule Syntax

History Limits

Suspend Scheduled Backups

Monitoring Backups

List All Backups

Check Backup Status

Metrics

Alerts

Restore from Backup

Create Cluster from Backup

Backup Workflow Details

Best Practices

Backup Timing

Retention Strategy

Disaster Recovery

S3 Bucket Configuration

Troubleshooting

Backup Stuck in Pending

Backup Failed with S3 Error

Restore Not Working

Build docs developers (and LLMs) love

Get Started

Core Concepts

Configuration

Operations

Runbooks

​Backup Methods

​RDB (Redis Database)

​AOF (Append-Only File)

​On-Demand Backups

​Basic Backup

​Backup Targets

​S3-Compatible Storage

​Scheduled Backups

​Daily Backup Schedule

​Schedule Syntax

​History Limits

​Suspend Scheduled Backups

​Monitoring Backups

​List All Backups

​Check Backup Status

​Metrics

​Alerts

​Restore from Backup

​Create Cluster from Backup

​Backup Workflow Details

​Best Practices

​Backup Timing

​Retention Strategy

​Disaster Recovery

​S3 Bucket Configuration

​Troubleshooting

​Backup Stuck in Pending

​Backup Failed with S3 Error

​Restore Not Working

Build docs developers (and LLMs) love

Backup Methods

RDB (Redis Database)

AOF (Append-Only File)

On-Demand Backups

Basic Backup

Backup Targets

S3-Compatible Storage

Scheduled Backups

Daily Backup Schedule

Schedule Syntax

History Limits

Suspend Scheduled Backups

Monitoring Backups

List All Backups

Check Backup Status

Metrics

Alerts

Restore from Backup

Create Cluster from Backup

Backup Workflow Details

Best Practices

Backup Timing

Retention Strategy

Disaster Recovery

S3 Bucket Configuration

Troubleshooting

Backup Stuck in Pending

Backup Failed with S3 Error

Restore Not Working