DataSync, Storage Gateway, and getting data in

Lesson 4 of 4 in Data Architecture for ML: Lakes, Warehouses, Catalogs.

Before any of this architecture matters, the data has to arrive. Four movers cover the exam's scenarios. AWS DataSync — online transfer of file and object data (NFS, SMB, HDFS, other clouds) into S3: scheduled, incremental, verified, bandwidth-throttled; the answer for migrations and recurring bulk syncs. Storage Gateway (File Gateway) — for applications that must keep writing to an NFS/SMB share while the files durably land in S3; ongoing hybrid access, not a one-way move. AWS Transfer Family — managed SFTP/FTPS/FTP endpoints that write straight into S3, so partners keep their existing upload clients. And when the pipe itself is the bottleneck, the Snow family in one line: if the network cannot carry it in time, ship a device.

Databases ride separately (AWS DMS replicates them continuously), and the recurring decision is Kinesis vs DataSync: a continuous stream of records being born versus periodic movement of files that already exist.

Interactive sorting exercise: Pick the mover for each scenario.

Interactive flashcard deck.

Tool: MLA-C02 mock exam — You have finished the heaviest domain on the blueprint. Run a Data Preparation domain drill in the Exam Hall — 28% of the real exam samples from exactly what these four modules taught.

Interactive checkpoint quiz (2 questions) — open this page in a browser to take it.