Lesson 09 — Amazon EKS Backup & Disaster Recovery

Learning Objectives

By the end of this lesson, you will be able to:

Understand backup and disaster recovery concepts
Explain business continuity principles
Identify Amazon EKS components that require backup
Design enterprise backup strategies
Implement Velero for Kubernetes backups
Protect persistent volumes
Understand AWS Backup integration
Design multi-AZ and multi-region recovery
Build ransomware-resilient backup architectures
Develop disaster recovery runbooks
Define Recovery Point Objectives (RPO) and Recovery Time Objectives (RTO)
Build enterprise backup governance and compliance

Why Backup & Disaster Recovery Matter

Security is not only about preventing attacks.

Organizations must prepare for:

Human error
Infrastructure failures
Ransomware
Region failures
Accidental deletion
Insider threats
Kubernetes misconfiguration
Cloud outages
Data corruption

A production Amazon EKS environment without a tested recovery strategy can experience prolonged outages and permanent data loss.

Failure

↓

Detection

↓

Recovery

↓

Business Continuity

What is Backup?

A backup is a protected copy of data that can be restored after loss or corruption.

Backups help recover from:

Deleted Kubernetes resources
Corrupted databases
Deleted namespaces
Storage failures
Malware
Ransomware
Configuration mistakes

What is Disaster Recovery?

Disaster Recovery (DR) is the process of restoring IT services after a major disruption.

A disaster may affect:

Cluster
AWS Account
Availability Zone
Region
Storage
Network
Identity services

DR includes:

Recovery planning
Automation
Testing
Documentation
Recovery validation

Business Continuity

Business Continuity ensures critical business services remain available even during major incidents.

Incident

↓

Recovery

↓

Business Continues

Business Continuity is broader than backups.

Backup vs Disaster Recovery

Backup	Disaster Recovery
Protects data	Restores services
Focuses on copies	Focuses on operations
Usually automated	Usually involves runbooks
Recover files/resources	Recover complete environments

Amazon EKS Components

Everything in Amazon EKS does not require the same backup strategy.

Amazon EKS

├── Kubernetes Resources
├── Namespaces
├── Deployments
├── StatefulSets
├── ConfigMaps
├── Secrets
├── Persistent Volumes
├── Amazon EBS
├── Amazon EFS
├── Application Data
├── Container Images
└── Infrastructure as Code

What Should Be Backed Up?

Critical resources include:

Namespaces
Deployments
StatefulSets
DaemonSets
Services
Ingress
ConfigMaps
Secrets
PVCs
CRDs
RBAC
Network Policies
Storage
Velero metadata

What Should NOT Be Backed Up?

Generally unnecessary:

ReplicaSets
Running Pods
Temporary Jobs
Cached data
Metrics
Ephemeral storage

These are recreated automatically.

Backup Architecture

Amazon EKS

↓

Velero

↓

Amazon S3

↓

EBS Snapshots

↓

Recovery

Types of Backups

Enterprise backup strategies include:

Full Backup
Incremental Backup
Differential Backup
Snapshot Backup
Application Backup
Database Backup

Recovery Objectives

Every workload should define:

Recovery Point Objective (RPO)

Maximum acceptable data loss.

Example:

15 Minutes

Recovery Time Objective (RTO)

Maximum acceptable recovery time.

Example:

30 Minutes

Example

Payment Platform

RPO:

5 Minutes

RTO:

30 Minutes

Development Cluster

RPO:

24 Hours

RTO:

8 Hours

Backup Frequency

Workload	Frequency
Critical	Hourly
Production	Daily
Non-production	Daily
Development	Weekly

Velero

Velero is the de facto Kubernetes backup and restore solution.

Velero backs up:

Kubernetes resources
Persistent Volumes
Snapshots
Metadata

Velero Architecture

Amazon EKS

↓

Velero Server

↓

S3 Bucket

↓

Snapshots

↓

Restore

Velero Components

Velero

├── Server
├── CLI
├── Backup Storage
├── Snapshot Plugins
└── Restore Controller

Installing Velero

Example:

velero install \
--provider aws \
--plugins velero/velero-plugin-for-aws

Creating a Backup

velero backup create production-backup

Listing Backups

velero backup get

Restoring

velero restore create \
--from-backup production-backup

Namespace Backup

velero backup create payments \
--include-namespaces payments

Persistent Volume Backup

Persistent Volumes require storage-level protection.

Amazon EBS volumes are backed up using snapshots.

PVC

↓

PV

↓

EBS Volume

↓

EBS Snapshot

Amazon EBS Snapshots

Benefits:

Incremental
Fast
Managed
Encrypted
Cross-region copy
Lifecycle policies

Amazon EFS Backup

Amazon EFS integrates with:

AWS Backup
Native backups
Cross-region replication

Suitable for:

Shared storage
Kubernetes shared volumes

AWS Backup

AWS Backup centralizes backups across AWS services.

Supports:

EBS
EFS
RDS
DynamoDB
FSx
EC2

AWS Backup Architecture

AWS Resources

↓

AWS Backup

↓

Backup Vault

↓

Recovery

Backup Vault

Backup Vaults provide:

Encryption
IAM protection
Retention
Access control
Compliance

Immutable Backups

Immutable backups cannot be modified or deleted before retention expires.

Benefits:

Ransomware protection
Insider protection
Compliance

Ransomware Protection

Enterprise backups should be:

Offline
Immutable
Encrypted
Versioned
Monitored

Encryption

Backups should always be encrypted.

Use:

AWS KMS
Customer-managed keys

Cross-Region Backup

Primary Region

↓

Backup Copy

↓

Secondary Region

Protects against regional disasters.

Multi-AZ Recovery

AZ-A

↓

AZ-B

↓

AZ-C

Supports high availability.

Multi-Region Disaster Recovery

Primary Region

↓

Backup Replication

↓

Secondary Region

↓

Recovery

Disaster Recovery Strategies

Strategy	Recovery Time
Backup & Restore	Hours
Pilot Light	Tens of minutes
Warm Standby	Minutes
Active-Active	Near Zero

Backup Testing

Never assume backups work.

Regularly test:

Restore
Integrity
Application startup
Database consistency

Disaster Recovery Testing

Exercise:

Cluster failure
Namespace recovery
Database recovery
Region recovery
Account recovery

Recovery Workflow

Incident

↓

Assessment

↓

Restore

↓

Validation

↓

Application Testing

↓

Production

Infrastructure as Code

Infrastructure should be recreated from:

Terraform
CloudFormation
GitOps

Infrastructure should not depend solely on backups.

GitOps

Git should remain the source of truth.

Git

↓

Terraform

↓

Amazon EKS

↓

Applications

Container Images

Images should be stored in:

Amazon ECR

Do not rely on running Pods.

Secrets Recovery

Secrets should come from:

AWS Secrets Manager
Parameter Store

Avoid storing production secrets only in Kubernetes.

Backup Security

Protect:

Backup bucket
Backup vault
IAM roles
Encryption keys
Restore permissions

Least Privilege

Only authorized users should:

Create backups
Delete backups
Restore backups

Monitoring

Monitor:

Backup failures
Snapshot failures
Storage usage
Restore failures
Expired backups

Compliance

Maintain evidence of:

Backup schedules
Restore testing
Encryption
Retention
DR exercises

Backup Lifecycle

Create

↓

Verify

↓

Encrypt

↓

Replicate

↓

Retain

↓

Expire

Backup Retention

Example:

Backup	Retention
Daily	30 Days
Weekly	3 Months
Monthly	1 Year
Annual	7 Years

Recovery Validation

After restore verify:

Pods Running
Services Healthy
Secrets Available
Storage Mounted
Database Connected

Common Backup Mistakes

Examples include:

Never testing restores
Single-region backups
Unencrypted backups
No retention policy
Missing IAM protection
No application validation
Missing Secrets
Backuping ephemeral resources

Enterprise Architecture

Amazon EKS

↓

Velero

↓

AWS Backup

↓

Amazon S3

↓

Cross-Region Copy

↓

Immutable Vault

↓

Recovery

Enterprise Implementation Strategy

Phase 1

Inventory workloads.

Phase 2

Define RPO and RTO.

Phase 3

Implement Velero.

Phase 4

Configure AWS Backup.

Phase 5

Enable encryption.

Phase 6

Enable cross-region replication.

Phase 7

Perform restore testing.

Phase 8

Automate DR runbooks.

Phase 9

Train operations teams.

Phase 10

Continuously improve.

Enterprise Best Practices

As a Cloud Security Engineer:

Backup Kubernetes resources.
Protect persistent volumes.
Use AWS Backup where appropriate.
Use Velero for Kubernetes recovery.
Encrypt all backups.
Store backups in multiple regions.
Use immutable backup vaults.
Test restores regularly.
Protect backup IAM roles.
Monitor backup failures.
Define RPO and RTO.
Practice disaster recovery exercises.
Store infrastructure as code in Git.
Protect Secrets using AWS Secrets Manager.
Document every recovery procedure.

Real-World Scenario

A multinational bank operates more than 250 Amazon EKS clusters across multiple AWS Regions.

A ransomware attack encrypts several worker nodes and deletes multiple namespaces.

Fortunately:

Velero has been performing scheduled Kubernetes backups to Amazon S3.
Amazon EBS snapshots protect persistent volumes.
AWS Backup stores encrypted copies in immutable Backup Vaults.
Cross-Region backup replication provides an off-site copy.
Infrastructure is recreated using Terraform.
GitOps restores Kubernetes manifests.
Secrets are retrieved from AWS Secrets Manager.
Application databases are recovered from recent snapshots.

The incident response team:

Isolates the affected clusters.
Rebuilds compromised worker nodes using approved AMIs.
Restores Kubernetes resources with Velero.
Restores EBS volumes from snapshots.
Recovers Secrets through AWS Secrets Manager.
Validates application integrity and business functionality.
Rotates any potentially exposed credentials.
Conducts a post-incident review and updates backup and recovery runbooks.

The result is that production services are restored within the defined RTO, data loss remains within the approved RPO, and the organization resumes normal operations with minimal business impact.

Key Takeaways

Backups protect data, while disaster recovery restores business services.
Define Recovery Point Objectives (RPO) and Recovery Time Objectives (RTO) for every workload.
Velero is the standard Kubernetes backup and restore solution for Amazon EKS.
Amazon EBS snapshots and AWS Backup protect persistent storage.
Store backups in encrypted, immutable and cross-region locations.
Test backup restoration regularly rather than assuming backups work.
Infrastructure as Code and GitOps accelerate recovery and reduce configuration drift.
Secrets should be recovered from centralized secret-management services such as AWS Secrets Manager.
Disaster recovery plans, runbooks and recovery exercises are as important as the backups themselves.

Knowledge Check

1. What is the difference between a backup and disaster recovery?

Answer: A backup is a copy of data for recovery, while disaster recovery is the end-to-end process of restoring applications, infrastructure and business services after a disruption.

2. What are RPO and RTO?

Answer: RPO defines the maximum acceptable amount of data loss, while RTO defines the maximum acceptable time to restore services after an outage.

3. What is Velero used for in Amazon EKS?

Answer: Velero is used to back up and restore Kubernetes resources, metadata and persistent volumes in Amazon EKS.

4. Why are immutable backups important?

Answer: Immutable backups cannot be altered or deleted before their retention period expires, providing strong protection against ransomware, insider threats and accidental deletion.

5. Why should disaster recovery plans be tested regularly?

Answer: Regular testing verifies that backups are recoverable, runbooks are accurate, recovery objectives can be met and teams are prepared to respond effectively during real incidents.

What’s Next?

In the next lesson, we will explore Lesson 10 — Enterprise Amazon EKS Security Architecture, where we will combine identity, networking, runtime security, logging, governance, backup and compliance into a complete production-ready Amazon EKS security reference architecture used by large enterprises.