Data Transfer & Hybrid Storage — Snowball, DataSync, Storage Gateway

The 4 tools

Tool	What	For
Snowball Edge	Physical device (80TB) shipped to AWS	HUGE data, slow internet (100TB over 100Mbps = 100 days; box = 1 week)
Snowmobile	Truck of storage (100PB)	Exabyte-scale migration
DataSync	Online transfer agent (rsync-like)	One-time migration + continuous sync, incremental
Storage Gateway	Hybrid bridge — local storage with cloud backend	On-prem apps keep running, data backed up to AWS automatically

Mental models

Snowball   = ship a hard drive box (one-time big move)
Snowmobile = ship a truck (petabyte scale)
DataSync   = sync over internet (ongoing / migration)
Gateway    = cloud is the HIDDEN backup of local storage

Storage Gateway modes

• File Gateway: NFS/SMB share → backed by S3
• Volume Gateway: iSCSI block → backed by EBS snapshots
• Tape Gateway: virtual tape → backed by Glacier

Storage Gateway = off-site backup (the core use case)

On-prem app keeps LOCAL storage (fast, low latency, apps unchanged)
Data continuously synced to S3/Glacier (automatic, no manual uploads)

On-prem dies / ransomware / office fire
  → restore from the AWS copy

Covers: file loss (S3 versioning), whole server death (rebuild from volume), office destroyed (data safe, rebuild from cloud).

Decision

Move once, huge, slow internet   → Snowball (box) / Snowmobile (truck)
Migrate or sync online           → DataSync
On-prem app needs cloud backend  → Storage Gateway (bridge)
Fast uploads from web users      → S3 Transfer Acceleration

vs AZs (not the same)

• AZs = AWS-internal high availability (survive building failure inside AWS)
• Storage Gateway = hybrid (connect your ON-PREM to AWS, off-site backup)
• Both = durability, but different worlds: on-prem ↔ AWS-internal

Exam facts

• “200TB over slow internet” → Snowball
• “Petabytes” → Snowmobile
• “sync on-prem to S3 continuously” → DataSync
• “on-prem app + cloud backup backend” → Storage Gateway
• “restore after on-prem failure” → Storage Gateway (off-site backup)

Related

• [[aws/index]]
• [[aws/storage-types/index]]
• [[aws/global-infrastructure/index]]