Amazon S3 Glacier
In-memory vault + archive store. Archives are uploaded as a single shot (multipart upload is not implemented). The job lifecycle is fast-forwarded — InitiateJob returns a job that is already StatusCode: Succeeded so callers don't have to poll.
Endpoint: http://localhost:4566Signing name: glacierProtocol: REST-JSON
Operations
| Operation | Method / Path |
|---|---|
CreateVault | PUT /{accountId}/vaults/{vaultName} |
DescribeVault / ListVaults / DeleteVault | /{accountId}/vaults[...] |
UploadArchive | POST /{accountId}/vaults/{vaultName}/archives |
DeleteArchive | DELETE /{accountId}/vaults/{vaultName}/archives/{archiveId} |
InitiateJob / DescribeJob / ListJobs | /{accountId}/vaults/{vaultName}/jobs[...] |
SetVaultNotifications / GetVaultNotifications / DeleteVaultNotifications | /{accountId}/vaults/{vaultName}/notification-configuration |
Behavior notes
DeleteVaultis rejected withInvalidParameterValueExceptionwhen the vault still contains archives.UploadArchiveaccepts the body as a base64-encoded blob; the response includes the SHA-256 of the bytes (Glacier's tree-hash collapses to a flat hash here).InitiateJobechoes backType,ArchiveId,Description,SNSTopic, andTierfromjobParameters, then immediately marks the jobSucceeded/Completed.Vault.NumberOfArchivesandVault.SizeInBytesare kept in sync as archives are uploaded and deleted.