Search the docs

When to Use GIR

GIR is appropriate when you need to:

  • Retain large volumes of data for compliance, archival, or reference workflows

  • Keep that data immediately accessible (millisecond reads) without restoring from a deeper archive tier

  • Reduce monthly storage costs significantly compared to S3 Standard

Typical workloads include media asset libraries, healthcare information archiving, legal information archiving, scientific datasets, and any long-lived, infrequently accessed data that must remain instantly retrievable.

GIR is not appropriate for:

  • Active working sets that are frequently read or written

  • Data that is read often enough that egress costs would exceed the storage savings

  • Files that are very small (under 128 KB) in large quantities — AWS bills each object at a minimum of 128 KB

  • Data with a short lifecycle — AWS charges a minimum of 90 days of storage per object regardless of when it is deleted

  • Data used as a replication transfer conduit between Hammerspace sites