Replicating MySQL to Redshift via Binlog CDC and Glue, Without DMS

Nightly full-dump pipelines to Redshift cannot capture updates or deletes as they happen, which leaves reports stale. The article compares common approaches: export-and-COPY, AWS DMS, zero-ETL integration, and managed ELT tools. It then presents a tutorial using OLake Go, an open-source tool that reads the MySQL binary log and writes Apache Iceberg tables into the AWS Glue Data Catalog. Redshift queries these tables in place through an external schema, and the author reports running the steps on a MySQL 8.4 and Redshift Serverless setup in ap-southeast-1.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.



Discussion (0)
Log in to join the discussion and vote.
Log in