9.1 Primer on Cloud Storage
- Data classification: Categorise data by type, sensitivity, and criticality to apply appropriate security controls and align with compliance strategies.
- Types of cloud storage:
- Object storage: Stores files as objects with metadata and unique IDs. Highly scalable, ideal for unstructured data. Provider manages redundancy; customer manages governance, backups, encryption.
- Volume/block storage: Acts as virtual hard drives attached to workloads. Customer manages redundancy, encryption, backups.
- Database storage: Managed relational (SQL) and non-relational (NoSQL) databases. Providers offer services like Amazon RDS, Google Cloud SQL, Azure SQL, DynamoDB, Cosmos DB.
- Other storage: Includes logging services, message queues, caches, in-memory databases, and SaaS storage (e.g., Google Drive, OneDrive).
9.2 Data Security Tools and Techniques
- Data classification: Categorises data by sensitivity and impact, guiding protection and compliance strategies. Requires continuous evaluation and clear ownership.
- Identity and Access Management (IAM): Governs access to resources via policies, both user-based and resource-based.
- Access policies: Define permissions and network rules to enforce security boundaries.
- Encryption and key management: Protects data by converting it to ciphertext; keys must be securely stored and managed, ideally separate from the CSP.
- Data Loss Prevention (DLP): Identifies, monitors, and protects sensitive data from unauthorised sharing or exfiltration. More common for SaaS due to cloud scale challenges. [D9 | PDF]
9.3 Cloud Data Encryption at Rest
- Encryption layers: Data can be encrypted at volume/object, file/API, database, or application layer. Higher layers offer more granular control but greater complexity.
- Application-level encryption: Encrypts sensitive data before database storage; protects data even from DB admins.
- File/API encryption: Granular protection for specific files or API-accessed data.
- Database encryption: Secures entire databases or specific tables/columns; often uses transparent database encryption (TDE).
- Object storage encryption: Automatic, provider-managed encryption for services like S3, Blob Storage; meets compliance standards.
- Volume encryption: Secures virtual disks; seamless and automated, protects data at rest and in backups/snapshots.
9.3.2 Cloud Data Key Management Strategies
- Client-side encryption: Customer encrypts data before uploading; provider never sees plaintext.
- Server-side encryption: Provider encrypts data using its own keys; easy to set up.
- Customer-managed encryption keys: Customer controls key lifecycle via provider’s KMS.
- Customer-provided encryption keys (BYOK): Customer creates and manages master keys; offers more control but greater responsibility.
- Custom application-level encryption: Hybrid scenarios where customer manages both encryption and keys.
9.3.3 Data Encryption Recommendations
- Use provider KMS for key management.
- For SaaS, rely on provider’s encryption tools; some allow customer-managed keys.
- Default encryption is often sufficient for compliance.
- Use different keys for different services/deployments.
- Apply IAM policies to keys for least privilege.
- Align encryption strategies with threat models (e.g., protect against credential compromise).
9.4 Data Security Posture Management (DSPM)
- DSPM tools focus on data-centred security: discovery, classification, access control evaluation, and remediation.
- DSPM helps visualise who has access to data and how, offering recommendations and managing overlapping controls.
9.5 Object Storage Security
- Object storage (e.g., S3, Blob Storage) poses exposure risks due to misconfigurations and complex access settings.
- Providers offer tools to block public access; encryption with KMS adds security.
- CDNs can enable safe public access to private storage.
- Continuous monitoring with CSPM and DSPM is essential.
9.6 Data Security for Artificial Intelligence
- AI systems require robust data security for algorithms and data assets.
- AI as a Service (AIaaS): Providers offer AI capabilities via subscription (e.g., Claude, ChatGPT, Vertex AI).
- Key considerations:
- Data deletion/retention policies
- Data flow understanding
- Provider’s security measures against adversarial attacks
- SLAs, security practices, regulatory compliance
Flashcards: https://quizlet.com/in/1125655369/ccsk-domain-9-data-security-flash-cards/?i=4jehw4&x=1jqt
Discover more from Information Security Blogs
Subscribe to get the latest posts sent to your email.
