Question: How to import data from S3 to DynamoDB?

Answer

Use DynamoDB's Import from S3 feature to load CSV, DynamoDB JSON or Amazon Ion into a new table. The managed import does not load into an existing table and does not consume that new table's write capacity.

1. Put the input files in S3 under a dedicated prefix. Choose a supported format and compression: uncompressed, GZIP or ZSTD.
2. Start an import from the DynamoDB console or ImportTable API, specifying the bucket, prefix, format and target table definition.
3. Define the new table's primary key and any GSIs, then check that every source item has correctly typed key attributes. LSIs are not supported by this import feature.
4. Monitor the import status and error details. Validate item counts and sample records before switching application traffic.

Plain application JSON is not automatically DynamoDB JSON; use the JSON converter to understand the typed representation. Duplicate primary keys can overwrite one another during import, so reconcile the input rather than assuming each source record creates a distinct item.

For an existing table, read and transform the files in an application or ETL job and write them through the normal APIs, with capacity and retry handling. Dynobase's S3 import workflow helps configure supported imports.

Import charges depend on uncompressed source data processed, including records that fail validation. S3 and related charges can also apply.

Source: AWS Import from S3.