Why can't AWS Lambda replace AWS Glue for larger workloads?

Lambda is designed for short, stateless, event‑driven functions with a maximum 15‑minute runtime and 10 GB memory. Glue, by contrast, is a managed ETL service that provides job scheduling, a data catalog, automatic schema discovery, and the ability to run long‑running, distributed transformations that can process petabytes of data. Glue also handles incremental data loading, job retries, and provides a visual interface for building ETL jobs. Because Lambda lacks these built‑in features, it cannot efficiently replace Glue for complex, large‑scale data pipelines that require orchestration, long runtimes, and heavy data movement.