Reviewed by Aditya Kumar · Last reviewed 2026-03-24
**S3**: Object storage. Key-value; eventual consistency (strong for new overwrites). No filesystem. Use for data lake, backup, static assets. Cheaper at scale ($0.023/GB). **EFS**: File system (NFS). Shared across EC2/ECS/Lambda. Use when apps need POSIX, shared storage, or...
S3: Object storage. Key-value; eventual consistency (strong for new overwrites). No filesystem. Use for data lake, backup, static assets. Cheaper at scale ($0.023/GB). EFS: File system (NFS). Shared across EC2/ECS/Lambda. Use when apps need POSIX, shared storage, or random write. Trade-off: EFS is more expensive; S3 is cheaper and scales infinitely. EFS when you need mount point, multi-writer, or legacy app compatibility. Use S3: Data lakes, ETL outputs, backups. Use EFS: Shared config, ML checkpoints, multi-container persistence. Cost: EFS ~$0.30/GB; S3 ~$0.023. At 1 TB, EFS ~$300 vs S3 ~$23.
Pro-Move: 'We use S3 for all pipeline data; EFS only for our ML training checkpoint sharing across 20 containers.' Red Flag: Using EFS for data lake—wrong tool; use S3.
According to DataEngPrep.tech, this is one of the most frequently asked Cloud/Tools interview questions, reported at 1 company. DataEngPrep.tech maintains an editor-reviewed database of 1,863 data engineering interview questions across 7 categories.