AWS announced managed migration from Apache Hive to Apache Iceberg for Amazon S3 Tables on November 4, 2026. The path is available in the Regions AWS names and accepts Hive tables stored as Parquet in Amazon S3, registered in AWS Glue, with the source and destination in the same AWS account and Region. The migration creates a new S3 Tables destination rather than converting the source in place, according to AWS.
Which Hive tables qualify
At launch, a source must be a Hive table registered in AWS Glue and stored as Parquet in an S3 general purpose bucket. The source and destination must be in the same account and Region. That rules out using this managed path for a source in a different account or Region.
How the migration proceeds
AWS’s console offers a “Create table from migration” entry point. The setup form asks for the Glue table source and destination details. The screenshots show the console interface, not a customer migration result.
AWS says the job checks compatibility, validates data, maps Hive types and partitioning to Iceberg, and handles copying, compression or recompression, and compaction. Teams can provide a destination schema or use one generated from the Glue-registered schema; migration policies govern mismatches. The destination is a new managed Iceberg table, and it stays read-only while migration is underway.
AWS’s migration documentation says the job proceeds automatically after planning by default. Operators can pause it to inspect compatibility reports before copying begins. After copying, an object-migration report gives per-object results, and progress statistics help track the job. AWS’s migration guide describes these review and reporting controls.
Syncs, writes, and cutover
The source remains available for reads and writes during the migration job, AWS says. Teams can run repeated syncs before completion; each copies changes, reducing the remaining work at cutover. AWS says source writes are interrupted for a few minutes during that final step. Once migration is complete, the destination becomes read-write.
AWS names Amazon Athena and Amazon SageMaker Unified Studio as integrations for working with S3 Tables, and says tools compatible with the Iceberg REST API can also connect.
US Regions and migration charges
AWS lists US East (N. Virginia, Ohio) and US West (Oregon) among the launch Regions. Because the source and destination must be in the same Region, a migration must meet that Region requirement as well as the same-account condition.
AWS says it does not charge for Hive-to-Iceberg conversion; customers pay for data copied. Objects in S3 Intelligent-Tiering remain in their current access tier during migration, and their days-since-last-access count is not reset.
