Study Guide: Selecting Appropriate Application Transfer Mechanisms

Learning Objectives

After studying this guide, you should be able to:

Distinguish between online and offline data transfer mechanisms based on volume and bandwidth constraints.
Select the correct AWS service for migrating specific workloads (databases, servers, or raw files).
Evaluate the cost-effectiveness of the AWS Snow Family versus online transfer methods.
Design a migration path using the Assess, Mobilize, Migrate, and Modernize phases.
Identify the role of AWS MGN and AWS DMS in an application migration strategy.

Key Terms & Glossary

AWS MGN (Application Migration Service): The primary service for lift-and-shift migrations, replicating source servers into AWS.
AWS DMS (Database Migration Service): A service that helps migrate databases to AWS quickly and securely while the source database remains functional.
AWS SCT (Schema Conversion Tool): Used to convert a source database schema to a format compatible with the target AWS database (e.g., Oracle to Aurora).
AWS DataSync: An online data transfer service that simplifies, automates, and accelerates moving data between on-premises storage and AWS S3/EFS/FSx.
AWS Snow Family: A collection of physical devices (Snowcone, Snowball, Snowmobile) used to migrate data physically from on-premises to AWS.

The "Big Idea"

Migration to the cloud is not a "one size fits all" process. It follows a lifecycle (Assess $\rightarrow$ Mobilize $\rightarrow$ Migrate $\rightarrow Modernize). The challenge for a Solutions Architect is to balance Volume (how much data?), Velocity (how fast must it move?), and Variety (is it a server image, a database, or a raw file?) to choose the most cost-effective and secure mechanism.

Formula / Concept Box

Concept	Metric / Rule
Transfer Time	T = \frac{\text{Total Data Volume}}{\text{Available Bandwidth} \times \text{Efficiency Factor}}
Snowball Rule	Usually not cost-effective if total data is < 10 TB.
Snowmobile Rule	Recommended for Exabyte-scale or multiple Petabytes (> 10$ PB).
S3 Transfer Acceleration	Best for geographically dispersed users uploading small-to-medium files to S3.

Hierarchical Outline

Data Transfer Strategies
- Online Transfers (Network-based)
  - AWS DataSync: For active/ongoing migrations and sync.
  - AWS Transfer Family: Managed SFTP, AS2, FTPS, and FTP.
  - Amazon Kinesis Data Firehose: For continuous streaming data.
- Offline Transfers (Physical-based)
  - AWS Snowcone: 8TB-14TB (Small, rugged, portable).
  - AWS Snowball Edge: 80TB-100TB (Compute and storage optimized).
  - AWS Snowmobile: 100PB (40ft shipping container).
Database Migration
- Homogeneous: Same engine (e.g., MySQL to Aurora MySQL). Use DMS.
- Heterogeneous: Different engine (e.g., SQL Server to Aurora). Use SCT + DMS.
Server Migration
- AWS Application Migration Service (MGN): Block-level replication for rehosting.
- VMware Cloud on AWS: For hybrid vSphere environments.

Visual Anchors

Migration Decision Logic

Loading Diagram...

The Migration Lifecycle

\begin{tikzpicture}[node distance=2cm, every node/.style={rectangle, draw, rounded corners, fill=blue!10, text centered, minimum width=3cm, minimum height=1cm}] \node (assess) {Assess}; \node (mobilize) [right of=assess, xshift=1.5cm] {Mobilize}; \node (migrate) [right of=mobilize, xshift=1.5cm] {Migrate}; \node (modernize) [right of=migrate, xshift=1.5cm] {Modernize};

code

\draw[->, thick] (assess) -- (mobilize);
\draw[->, thick] (mobilize) -- (migrate);
\draw[->, thick] (migrate) -- (modernize);

\node[draw=none, fill=none, below of=assess, yshift=1cm] {\scriptsize Inventory/TCO};
\node[draw=none, fill=none, below of=mobilize, yshift=1cm] {\scriptsize Planning/Pilot};
\node[draw=none, fill=none, below of=migrate, yshift=1cm] {\scriptsize Lift & Shift (MGN)};
\node[draw=none, fill=none, below of=modernize, yshift=1cm] {\scriptsize Cloud Native};

\end{tikzpicture}

Definition-Example Pairs

Rehosting (Lift-and-Shift): Moving an application as-is to the cloud.
- Example: Using AWS MGN to move a legacy Windows 2012 server to an EC2 instance without changing the code.
Replatforming (Lift-and-Reshape): Making minor optimizations without changing core architecture.
- Example: Moving a self-managed Oracle database on-premises to Amazon RDS for Oracle.
Refactoring: Re-architecting the application to use cloud-native features.
- Example: Breaking a monolithic Java app into microservices running on AWS Lambda.

Worked Examples

Scenario: The 50TB Dilemma

Problem: A company needs to migrate 50 TB of data to AWS. They have a 100 Mbps internet connection dedicated to migration. Should they use online or offline transfer?

Step 1: Calculate Online Time

Bandwidth = 100 Mbps $\approx$ 12.5 MB/s.
50 TB = 50,000,000 MB.
Time = $50,000,000 / 12.5 = 4,000,000 $seconds$ \approx 46 days (assuming 100% efficiency).

Step 2: Compare to Offline

AWS Snowball Edge shipping and processing usually takes 5-10 business days.

Conclusion: Use AWS Snowball Edge. It is significantly faster and avoids saturating the network for over a month.

Checkpoint Questions

What is the minimum recommended data volume to make AWS Snowball cost-effective?
Which tool should you use first if you are migrating a database from SQL Server to Amazon DynamoDB?
When would you choose AWS DataSync over AWS Transfer Family?
What is the primary difference between AWS Snowball and AWS Snowmobile?

[!TIP] Answers: 1. 10 TB. 2. AWS SCT (to convert schema). 3. DataSync is for automated synchronization; Transfer Family is for specific protocol-based access (SFTP). 4. Snowmobile is for Exabyte scale (>10$ PB), whereas Snowball is for Petabyte scale.

Muddy Points & Cross-Refs

MGN vs. DMS: Students often confuse these. Remember: MGN is for the whole server (OS + Apps); DMS is specifically for the data inside the database.
Storage Gateway vs. DataSync: Use Storage Gateway for hybrid cache/access (on-premises users need cloud storage). Use DataSync for one-way or two-way migration/syncing tasks.

Comparison Tables

Feature	AWS DataSync	AWS Snowball Edge	AWS Transfer Family
Mechanism	Online (Network)	Offline (Physical)	Online (Protocol)
Speed	Up to 10 Gbps	Physical Shipping	Network Dependent
Use Case	Ongoing data sync	One-time large migration	SFTP/FTP Interface
Security	TLS Encryption	KMS & Ruggedized Case	IAM & Encryption