DynamoDB Adds Bulk Put Items to Load S3 Data into Existing Tables

AWS has introduced a DynamoDB feature that loads large volumes of data from S3 directly into a table that already exists and serves production traffic. Previously, S3 imports could only create new tables, so teams had to build custom pipelines with Lambda, Step Functions or Glue and temporarily raise capacity. The import runs without consuming the table's read or write capacity, and newly written items are readable while the job is still in progress. Each record in an input file can carry its own condition, which allows a nightly sync to be re-run safely without overwriting newer application writes. The article notes that cancelled runs are not rolled back and that written records still appear in DynamoDB streams.