Amazon Glacier

Overview

CloudMock emulates Amazon S3 Glacier, supporting vault management, archive upload/deletion, job initiation and retrieval, vault lock (two-step InProgress → Locked), and tagging. Job lifecycle transitions from InProgress → Succeeded.

Supported Operations

Operation	Status	Notes
CreateVault	Supported	Creates a new vault
DescribeVault	Supported	Returns vault details and inventory info
ListVaults	Supported	Lists all vaults
DeleteVault	Supported	Fails if vault is non-empty
UploadArchive	Supported	Stores archive, returns ID in x-amz-archive-id header
DeleteArchive	Supported	Deletes archive by ID
InitiateJob	Supported	archive-retrieval or inventory-retrieval
DescribeJob	Supported	Returns job status and details
ListJobs	Supported	Lists jobs for a vault
GetJobOutput	Supported	Returns mock output for completed jobs
InitiateVaultLock	Supported	Creates InProgress lock, returns lockId in header
CompleteVaultLock	Supported	Transitions lock to Locked state
AbortVaultLock	Supported	Cancels InProgress vault lock
GetVaultLock	Supported	Returns lock policy and state
AddTagsToVault	Supported	Adds tags (POST /-/vaults/{name}/tags)
RemoveTagsFromVault	Supported	Removes tags by key
ListTagsForVault	Supported	Lists vault tags

REST API Routing

Glacier uses REST path-based routing, not JSON action names:

Method	Path	Operation
PUT	/-/vaults/{name}	CreateVault
GET	/-/vaults/{name}	DescribeVault
DELETE	/-/vaults/{name}	DeleteVault
GET	/-/vaults	ListVaults
POST	/-/vaults/{name}/archives	UploadArchive
DELETE	/-/vaults/{name}/archives/{id}	DeleteArchive
POST	/-/vaults/{name}/jobs	InitiateJob
GET	/-/vaults/{name}/jobs/{id}	DescribeJob
GET	/-/vaults/{name}/jobs/{id}/output	GetJobOutput
POST	/-/vaults/{name}/lock-policy	InitiateVaultLock
POST	/-/vaults/{name}/lock-policy/{id}	CompleteVaultLock
DELETE	/-/vaults/{name}/lock-policy	AbortVaultLock
GET	/-/vaults/{name}/lock-policy	GetVaultLock
GET	/-/vaults/{name}/tags	ListTagsForVault
POST	/-/vaults/{name}/tags	AddTagsToVault

Quick Start

import boto3

client = boto3.client('glacier',
    endpoint_url='http://localhost:4566',
    region_name='us-east-1',
    aws_access_key_id='test',
    aws_secret_access_key='test')

client.create_vault(vaultName='my-archives')

# Upload archive
with open('backup.tar.gz', 'rb') as f:
    response = client.upload_archive(
        vaultName='my-archives',
        body=f,
        archiveDescription='Monthly backup',
    )
archive_id = response['archiveId']

# Initiate inventory job
job = client.initiate_job(
    vaultName='my-archives',
    jobParameters={'Type': 'inventory-retrieval'},
)
print(job['jobId'])

Configuration

services:
  glacier:
    enabled: true

Known Differences from AWS

• Archives are stored in memory (not persisted)
• Job output is mock data, not real archive bytes
• Inventory dates and archive counts are simulated