Skip to main content

Scheduled Backups

Configure automated backup schedules using the KafkaBackup CRD's scheduling capabilities.

Cron Schedule Format​

The schedule field uses 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)
│ │ │ │ │
│ │ │ │ │
* * * * *

Common Schedules​

ScheduleCron ExpressionDescription
Every hour0 * * * *At minute 0 of every hour
Every 15 minutes*/15 * * * *Every 15 minutes
Every 6 hours0 */6 * * *At minute 0 past every 6th hour
Daily at midnight0 0 * * *Every day at 00:00
Daily at 2 AM0 2 * * *Every day at 02:00
Weekly (Sunday)0 0 * * 0Every Sunday at 00:00
Monthly0 0 1 * *First day of month at 00:00

Examples​

Hourly Backup​

apiVersion: kafka.oso.sh/v1alpha1
kind: KafkaBackup
metadata:
  name: hourly-backup
  namespace: kafka-backup
spec:
  schedule: "0 * * * *"  # Every hour at minute 0

  kafkaCluster:
    bootstrapServers:
      - kafka:9092

  topics:
    - orders
    - payments

  storage:
    storageType: s3
    s3:
      bucket: kafka-backups
      region: us-west-2
      prefix: hourly

  compression: zstd
  compressionLevel: 3

  retention:
    backups: 168  # Keep 7 days of hourly backups

Daily Backup (Off-Peak Hours)​

apiVersion: kafka.oso.sh/v1alpha1
kind: KafkaBackup
metadata:
  name: daily-backup
  namespace: kafka-backup
spec:
  schedule: "0 2 * * *"  # Daily at 2 AM

  kafkaCluster:
    bootstrapServers:
      - kafka:9092
    securityProtocol: SASL_SSL
    saslSecret:
      name: kafka-credentials
      mechanism: SCRAM-SHA-256

  topics:
    - "*"  # All topics
  excludeTopics:
    - "__*"
    - "_*"

  storage:
    storageType: s3
    s3:
      bucket: kafka-backups
      region: us-west-2
      prefix: daily

  compression: zstd
  compressionLevel: 6  # Higher compression for archival

  retention:
    backups: 90  # Keep 90 days of daily backups

Weekly Archival Backup​

apiVersion: kafka.oso.sh/v1alpha1
kind: KafkaBackup
metadata:
  name: weekly-archive
  namespace: kafka-backup
spec:
  schedule: "0 3 * * 0"  # Weekly on Sunday at 3 AM

  kafkaCluster:
    bootstrapServers:
      - kafka:9092

  topics:
    - "*"

  storage:
    storageType: s3
    s3:
      bucket: kafka-archives
      region: us-west-2
      prefix: weekly

  compression: zstd
  compressionLevel: 9  # Maximum compression

  retention:
    backups: 52  # Keep 1 year of weekly backups

Multi-Tier Backup Strategy​

Implement comprehensive backup coverage:

# Tier 1: Frequent backups for low RPO
---
apiVersion: kafka.oso.sh/v1alpha1
kind: KafkaBackup
metadata:
  name: tier1-frequent
spec:
  schedule: "*/15 * * * *"  # Every 15 minutes

  kafkaCluster:
    bootstrapServers:
      - kafka:9092

  topics:
    - orders  # Critical topics only
    - payments

  storage:
    storageType: s3
    s3:
      bucket: kafka-backups
      prefix: tier1-frequent

  compression: lz4  # Fast compression
  retention:
    backups: 96  # 24 hours of 15-minute backups

---
# Tier 2: Hourly backups for all topics
apiVersion: kafka.oso.sh/v1alpha1
kind: KafkaBackup
metadata:
  name: tier2-hourly
spec:
  schedule: "0 * * * *"  # Hourly

  kafkaCluster:
    bootstrapServers:
      - kafka:9092

  topics:
    - "*"
  excludeTopics:
    - "__*"

  storage:
    storageType: s3
    s3:
      bucket: kafka-backups
      prefix: tier2-hourly

  compression: zstd
  compressionLevel: 3
  retention:
    backups: 168  # 7 days

---
# Tier 3: Daily archival
apiVersion: kafka.oso.sh/v1alpha1
kind: KafkaBackup
metadata:
  name: tier3-daily
spec:
  schedule: "0 2 * * *"  # Daily at 2 AM

  kafkaCluster:
    bootstrapServers:
      - kafka:9092

  topics:
    - "*"

  storage:
    storageType: s3
    s3:
      bucket: kafka-archives
      prefix: tier3-daily

  compression: zstd
  compressionLevel: 9
  retention:
    backups: 365  # 1 year

One-Time Backup​

For manual/on-demand backups, omit the schedule:

apiVersion: kafka.oso.sh/v1alpha1
kind: KafkaBackup
metadata:
  name: manual-backup-20241201
spec:
  # No schedule - runs once immediately

  kafkaCluster:
    bootstrapServers:
      - kafka:9092

  topics:
    - orders
    - payments

  storage:
    storageType: s3
    s3:
      bucket: kafka-backups
      prefix: manual

Trigger Manual Backup on Schedule​

Force immediate execution of a scheduled backup:

# Add annotation to trigger backup
kubectl annotate kafkabackup hourly-backup \
  kafka.oso.sh/trigger-backup=$(date +%s) \
  --overwrite

Backup Window Management​

Avoid Peak Hours​

spec:
  schedule: "0 2-5 * * *"  # Between 2 AM and 5 AM only

Weekend-Only Archival​

spec:
  schedule: "0 3 * * 0,6"  # Saturday and Sunday at 3 AM

Business Hours Exclusion​

# Backup every hour except 9 AM to 5 PM
spec:
  schedule: "0 0-8,18-23 * * *"

Timezone Considerations​

Kubernetes CronJobs use UTC by default. Adjust your schedule accordingly:

# For 2 AM EST (UTC-5), schedule at 7 AM UTC
spec:
  schedule: "0 7 * * *"

Or use the CRON_TZ prefix (if supported):

spec:
  schedule: "CRON_TZ=America/New_York 0 2 * * *"

Retention Configuration​

By Count​

retention:
  backups: 168  # Keep last 168 backups

S3 Lifecycle (Supplementary)​

Configure S3 lifecycle for additional storage tiering:

{
  "Rules": [
    {
      "ID": "BackupLifecycle",
      "Status": "Enabled",
      "Filter": {"Prefix": "hourly/"},
      "Transitions": [
        {"Days": 30, "StorageClass": "STANDARD_IA"},
        {"Days": 90, "StorageClass": "GLACIER"}
      ],
      "Expiration": {"Days": 365}
    }
  ]
}

Monitoring Schedules​

Check Next Run Time​

kubectl get kafkabackup hourly-backup \
  -o jsonpath='{.status.nextScheduledTime}'

View Schedule History​

kubectl get kafkabackup hourly-backup \
  -o jsonpath='{.status.backupHistory}' | jq

Alert on Missed Backups​

# PrometheusRule
apiVersion: monitoring.coreos.com/v1
kind: PrometheusRule
metadata:
  name: backup-schedule-alerts
spec:
  groups:
    - name: kafka-backup-schedule
      rules:
        - alert: BackupScheduleMissed
          expr: |
            (time() - kafka_backup_last_successful_backup_timestamp) >
            (kafka_backup_schedule_interval_seconds * 2)
          for: 5m
          labels:
            severity: warning
          annotations:
            summary: "Backup schedule missed"
            description: "{{ $labels.backup_name }} hasn't run on schedule"

Best Practices​

  1. Stagger schedules - Don't run all backups at the same time
  2. Use retention - Prevent storage from growing unbounded
  3. Match frequency to RPO - More frequent = lower data loss
  4. Use compression levels wisely - Higher for archival, lower for frequent
  5. Monitor schedule adherence - Alert on missed backups
  6. Test restores - Regularly verify backups are usable

Next Steps​