Last updated on April 7, 2023
- Amazon S3 is a durable, secure, simple, and fast storage service, while Amazon S3 Glacier is used for archiving solutions.
- Use S3 if you need low latency or frequent access to your data. Use S3 Glacier for low storage cost, and you do not require millisecond access to your data.
- You have three retrieval options when it comes to Glacier, each varying in the cost and speed it retrieves an object for you. You retrieve data in milliseconds from S3.
- Both S3 and Glacier are designed for durability of 99.999999999% of objects across multiple Availability Zones.
- S3 and Glacier are designed for availability of 99.99%.
- S3 can be used to host static web content, while Glacier cannot.
- In S3, users create buckets. In Glacier, users create archives and vaults.
- You can store a virtually unlimited amount of data in both S3 and Glacier.
- A single Glacier archive can contain 40TB of data.
- S3 supports Versioning.
- You can run analytics and querying on S3.
- You can configure a lifecycle policy for your S3 objects to automatically transfer them to Glacier. You can also upload objects directly to either S3 or Glacier.
- S3 Standard-IA and One Zone-IA have a minimum capacity charge per object of 128KB. Glacier’s minimum is 40KB.
- Objects stored in S3 have a minimum storage duration of 30 days (except for S3 Standard). Objects that are archived to Glacier have a minimum 90 days of storage. Objects that are deleted, overwritten, or transitioned to a different storage class before the minimum duration will incur the normal usage charge plus a pro-rated request charge for the remainder of the minimum storage duration.
- Glacier has a per GB retrieval fee.
- You can transition objects from some S3 storage classes to another. Glacier objects can only be transitioned to the Glacier Deep Archive storage class.
- S3 (standard, intelligent-tiering, standard-IA, and one zone-IA) and Glacier are backed by an SLA.