Introducing self-managed Amazon S3 buckets for AWS Lambda function code

This article announces self-managed Amazon S3 buckets for AWS Lambda deployment packages, allowing Lambda to read code directly from customer-owned buckets instead of copying to AWS-managed storage.

• Eliminates 75 GB code storage quota pressure by not counting deployment packages toward account limits
• Provides full security and compliance control: encryption, access policies, lifecycle management, and audit trails
• Enables disaster recovery using S3 Cross-Region Replication to maintain fallback code copies
• Improves performance with faster function creation and updates by eliminating internal copy operations
• Set S3ObjectStorageMode to REFERENCE when creating or updating functions to enable the feature
• Requires S3 versioning and supports SSE-S3, SSE-KMS, and DSSE-KMS encryption
• Supports lifecycle policies to archive old versions to Glacier or delete non-current versions automatically
• Works across AWS accounts and regions with proper bucket policies and cross-account access configuration

Self-managed S3 buckets give organizations greater capacity, control, and compliance capabilities for Lambda deployments without changing function code or invocation patterns.