//
Search across all documentation pages
7 pages in this section.
How Glacier and Deep Archive trade storage cost for retrieval latency: why the cheapest bytes are the slowest to read back, and how retrieval tiers let you buy speed per request.
The everyday archival operations in boto3 and AWS SDK v3 - writing an object straight into a Glacier class, checking its class and restore state, restoring a cold object, and reading GLACIER_IR instantly.
The modern way to archive on AWS: write and restore objects in the GLACIER_IR, GLACIER, and DEEP_ARCHIVE storage classes through the ordinary S3 API in boto3 and AWS SDK v3, instead of the legacy Glacier Vault service.
Let S3 tier objects into GLACIER_IR, GLACIER, and DEEP_ARCHIVE automatically as they age, using PutBucketLifecycleConfiguration transitions in boto3 and AWS SDK v3 - no cron job, no manual re-tiering.
A checklist for cost-safe, correct archival storage from the SDK - choosing classes, designing retention, restoring predictably, controlling retrieval cost, and preferring S3 Glacier classes over the legacy Vault API - across boto3 and AWS SDK v3.