Hybrid & Offline Storage: Storage Gateway & Snow Family
For the SAA-C03 exam, understanding how to bridge on-premises environments with the AWS Cloud is critical. AWS provides two primary paths: Storage Gateway for continuous hybrid integration and the Snow Family for large-scale, offline data migration and edge computing.
The Analogy: The Bridge vs. The Moving Truck
Imagine your data is a collection of books.
- AWS Storage Gateway is like a high-speed conveyor belt (Bridge) connecting your local library to a massive AWS warehouse. Books move back and forth automatically as you need them.
- AWS Snow Family is like a heavy-duty armored moving truck. When you have 100,000 books to move at once, the conveyor belt is too slow. You pack the truck, drive it to the warehouse, and unload everything in one go.
Core Concepts & Well-Architected Lens
1. Operational Excellence & Reliability
AWS Storage Gateway provides low-latency access to data by caching frequently used data on-premises while storing the full dataset in Amazon S3. This ensures that even if the internet connection is throttled, local operations can continue using cached data.
2. Performance Efficiency
Snowball Edge devices provide local computing power (EC2-compatible) and storage. This allows for “Edge Computing,” where data is processed locally at a remote site (like a ship or a mine) before being shipped back to AWS.
Service Comparison Table
| Service Type | Storage Type | Primary Use Case | Protocol |
|---|---|---|---|
| S3 File Gateway | Object (S3) | Flat files, backups to S3, low-latency local cache. | NFS / SMB |
| FSx File Gateway | File (FSx Windows) | Windows File Server migration, AD integration. | SMB |
| Volume Gateway | Block (EBS Snapshots) | iSCSI targets, local disk expansion, DR. | iSCSI |
| Tape Gateway | Archive (Glacier) | Replacing physical tape libraries (VTL). | iSCSI-VTL |
Decision Matrix (If/Then)
- If you need to migrate 100TB of data and have limited bandwidth: Use Snowball Edge Storage Optimized.
- If you need local servers to see S3 buckets as a network drive: Use S3 File Gateway.
- If you need to run pre-processing on data in a location with no internet: Use Snowball Edge Compute Optimized.
- If you need to replace on-prem physical tapes for long-term compliance: Use Tape Gateway.
Exam Tips: Golden Nuggets
- Snowball vs. Snowcone: Snowcone is tiny (8TB), fits in a backpack, and can be powered by a battery. Snowball is for large migrations (up to 80TB usable).
- Cached vs. Stored Volumes: Cached keeps only active data local (saves space); Stored keeps the entire dataset local (lowest latency) and backs up to AWS.
- The “10 TB” Rule: Generally, if it takes more than a week to upload data over the wire, the exam will point you toward a Snow device.
- Data Security: All Snow family devices use KMS-managed encryption keys and are physically ruggedized with TPM chips.
Hybrid Storage Architecture
- Snowcone: 8TB, portable.
- Snowball Edge: 80TB+, Storage/Compute.
- S3 File Gateway: Hybrid S3 access.
- Using Snowball for < 5TB (Not cost-effective).
- Forgetting that Snowball is offline (one-time move).
- Assuming Storage Gateway works without internet.
- Cloud Migration: Snowball Edge.
- Hybrid Backup: S3 File Gateway.
- Edge AI/ML: Snowball Compute Opt.