AWS Database Security

AWS Database Security Across RDS, Aurora and Redshift

The AWS shared responsibility model leaves database security with you. DBsafe delivers it.

AWS hosts the world's largest concentration of enterprise databases: RDS, Aurora, Redshift, DynamoDB, and Neptune serve critical workloads across every industry. Each service has a distinct security model, and each creates specific compliance obligations for the organisations that use them. DBsafe provides unified AWS database security assessment, PII discovery, and compliance reporting across all AWS database services.

47%
of cloud data breaches involve an AWS-hosted database
Public
publicly accessible RDS instances without strong passwords are found in every AWS audit
IAM
IAM database authentication is the recommended approach but rarely configured
VPC
databases outside VPC or with public subnets are a consistent high-risk finding
The Challenge

AWS database services are secure by capability but not by default

AWS provides comprehensive database security capabilities: encryption at rest and in transit, IAM database authentication, VPC isolation, Security Groups, CloudTrail logging, and GuardDuty threat detection. But these capabilities require deliberate configuration — the defaults are not sufficient for production security or compliance.

The rapid pace of AWS adoption means databases are often provisioned with default settings optimised for getting started quickly, and security hardening is deferred and then forgotten. The result is a common pattern: older RDS instances with public accessibility, no minor version auto-upgrade, and SQL authentication that was never migrated to IAM.

The cost of inaction

Publicly accessible RDS without strong password: directly exploitable by internet scanning tools

No AWS CloudTrail database events: insufficient audit trail for PCI DSS and HIPAA requirements

Unencrypted RDS instance: fails PCI DSS Requirement 3 and HIPAA encryption requirements

Security group allowing 3306/5432/1521 from 0.0.0.0/0: database port accessible from the internet

The DBsafe Approach

How DBsafe secures AWS database deployments

01
01

Assess configuration across all AWS database services

DBsafe checks RDS, Aurora, Redshift, and DynamoDB configurations against AWS security best practices: public accessibility, security groups, encryption, backup, minor version upgrades, and IAM authentication.

02
02

Discover sensitive data in all database types

Scan RDS, Aurora, and Redshift for PII, cardholder data, and health information. DynamoDB and DocumentDB discovery covers NoSQL document structures where PII accumulates without schema-level visibility.

03
03

Monitor queries beyond CloudTrail

Supplement CloudTrail data event logging with query-level monitoring, anomaly detection, and tamper-evident audit storage that provides the coverage compliance frameworks require.

RDS security configuration essentials

Amazon RDS security configuration covers four primary areas: network isolation, authentication, encryption, and monitoring. Network isolation: RDS should be placed in private subnets within a VPC, with Security Groups restricting access to specific application security groups and corporate IP ranges. Public accessibility should be disabled for all production instances.

Authentication: IAM database authentication is preferred over username-password for applications running within AWS, as it eliminates long-lived database credentials. Master user passwords should be rotated using AWS Secrets Manager. Encryption: all RDS instances should have encryption at rest enabled; KMS CMK is preferred for sensitive workloads.

Amazon Aurora security considerations

Aurora shares RDS's security model but adds Aurora-specific considerations: cluster topology (multi-AZ, global database, serverless v2), reader endpoint access control, and IAM role-based access for Aurora Serverless via Data API.

Aurora Global Database creates cross-region data flows that have GDPR cross-border transfer implications for EU-based workloads. DBsafe's Aurora assessment includes region-specific configuration checks and flags cross-region replication as a data transfer compliance item.

Amazon Redshift security for analytics workloads

Redshift, as a data warehouse, often holds analytical copies of operational data including PII from multiple source systems. Redshift security configuration includes: cluster VPC placement, encrypted connections, column-level access control (Late Binding Views, Column-Level Access Control feature), row-level security through RLS policies (Redshift serverless and RA3), and Redshift Audit Logging to S3.

DBsafe scans Redshift clusters for PII, assesses their security configuration, and verifies that column and row-level access controls are implemented for sensitive columns.

DynamoDB security and compliance

DynamoDB presents different security challenges from relational databases: its schema-less document structure means PII may appear in any attribute of any item without schema-level signal. DynamoDB security controls include: IAM policy-based access control (no native user management), encryption at rest with KMS, VPC endpoint for private network access, and DynamoDB Streams for change data capture.

DBsafe's DynamoDB discovery scans item samples for PII within document structures, providing the data classification needed to apply appropriate access controls and compliance evidence.

AWS CloudTrail and database audit logging

AWS CloudTrail records API calls to AWS services including RDS, but does not capture the SQL queries executed against the database. For PCI DSS Requirement 10, HIPAA Audit Controls, and SOC 2 CC7, SQL-level query logging is required.

DBsafe supplements CloudTrail API logging with SQL-level query capture, providing the combined audit record that compliance frameworks require: management-plane events (CloudTrail) and data-plane events (DBsafe) in a unified, tamper-evident store.

AWS encryption key management with KMS

AWS KMS provides managed encryption keys for RDS, Aurora, and Redshift encryption at rest. Two key models are available: AWS managed keys (managed entirely by AWS, no customer control) and customer managed keys (CMK, with customer-controlled policies, rotation, and access).

For PCI DSS Requirement 3.6 (secure management of cryptographic keys) and HIPAA requirements for encryption key control, CMK is typically required. DBsafe's AWS assessment checks the key management model for all encrypted databases and flags AWS managed key usage where CMK is required by your compliance framework.

Platform Capabilities

Everything you need to address AWS database security — in one integrated platform.

Multi-Service Assessment

Configuration assessment for RDS, Aurora, Redshift, DynamoDB, and DocumentDB.

NoSQL and Relational PII Discovery

Find sensitive data in DynamoDB document structures and relational RDS tables.

IAM and Security Group Analysis

Identify over-privileged IAM roles and permissive Security Groups exposing database ports.

SQL-Level Query Monitoring

Supplement CloudTrail with SQL query capture for complete compliance audit coverage.

KMS Configuration Assessment

Verify encryption key management including CMK requirements for regulated workloads.

AWS Compliance Reports

GDPR, PCI DSS, HIPAA, and SOC 2 evidence reports for AWS database environments.

Frequently Asked Questions

Answers to the most common questions about AWS database security.

Is Amazon RDS secure by default?

No. Amazon RDS provides secure capabilities but not secure defaults. Public accessibility can be enabled, encryption at rest defaults to disabled for some database engines, minor version upgrades must be explicitly enabled, and IAM database authentication requires configuration. DBsafe's RDS assessment checks all of these settings and provides prioritised remediation guidance.

Does AWS CloudTrail provide sufficient database audit logging for PCI DSS?

No. CloudTrail records API calls to AWS services (management plane) but does not capture SQL queries (data plane). PCI DSS Requirement 10 requires logging of user access to cardholder data, which requires SQL-level logging. DBsafe provides SQL query capture that, combined with CloudTrail, delivers the complete audit coverage PCI DSS requires.

How does DBsafe discover PII in DynamoDB?

DBsafe samples DynamoDB items across all tables and analyses the attribute values within each item. For string attributes, PII detection patterns are applied. For map and list attributes, recursive analysis examines nested structures. Items containing PII are flagged with the specific attribute paths where personal data was found.

Can DBsafe assess Amazon Aurora Serverless?

Yes. DBsafe supports Aurora Serverless v1 and v2. For Aurora Serverless v2, standard driver connections are used. For Aurora Serverless v1, the RDS Data API is used for query execution and monitoring. Assessment coverage includes serverless-specific configurations including ACU scaling, pause behaviour, and Data API authentication.

How does DBsafe help with Redshift column-level access control?

DBsafe discovers PII in Redshift tables and verifies that appropriate column-level access controls (Late Binding Views or column-level privileges) are in place for sensitive columns. Where sensitive columns lack appropriate access controls, DBsafe generates findings with specific remediation recommendations for column-level security implementation.

Ready in 5 minutes

Secure your AWS database estate with confidence

DBsafe assesses, monitors, and reports on AWS database security continuously — giving you the visibility and evidence your compliance programme requires.