Envelope Encryption Pattern
What you'll learn: You'll discover how envelope encryption solves the performance and scale challenges of encrypting large amounts of data by using a two-tier key approach.
The Core Problem
Encrypting data directly with a master key (like a KMS key) works fine for small amounts of data, but becomes impractical at scale. Why? Master keys in KMS services have strict API rate limits and network latency. Imagine encrypting terabytes of data by sending every byte through KMS—it would be impossibly slow and expensive.
How Envelope Encryption Works
Envelope encryption uses a two-layer approach:
- Generate a data encryption key (DEK) locally—a random symmetric key
- Encrypt your actual data with this DEK (happens locally, very fast)
- Encrypt the DEK itself with your master key (KEK) via KMS
- Store both: the encrypted data and the encrypted DEK together
When you need to decrypt:
- Send the encrypted DEK to KMS to decrypt it with the master key
- Use the plaintext DEK locally to decrypt your data
- Discard the plaintext DEK when done
The Analogy
Think of it like a safe deposit box (your data) locked with a physical key (DEK). Instead of carrying that physical key to a remote vault every time you open the box, you keep it in your pocket wrapped in a special envelope that only the bank manager (KMS master key) can open. You only visit the bank manager once to unwrap the key, then use it locally as many times as needed.
Why This Matters
Envelope encryption lets you encrypt gigabytes or terabytes of data with one KMS API call instead of millions. The heavy cryptographic work happens locally at memory speed, while KMS only protects the small DEK.
Key Takeaway: Envelope encryption scales data protection by encrypting data locally with data keys (DEKs), then encrypting only those small keys with KMS master keys (KEKs)—giving you both performance and centralized key control.