Amazon officially launched its cloud service solution for long term archiving, Glacier, on 21st August 2012. Named Glacier as it is intended to be vast and slow moving, with an extremely low storage cost to match.
Data can be uploaded as a single file or as multiple files which are then stored in archives. There is no limit to the amount of data you can store but each archive is limited to 40 Terabytes. There are no up-front costs and you only pay for what you store. Glacier will cost just $0.01 per Gigabyte per month compared to Amazons other storage offering, S3, which can cost up to $0.13 per and Googles Cloud Storage at $0.10, both at per Gigabyte per month. In theory using Glacier one could store 10GB of data over 10 years for just $1.20 providing you don’t access it. Amazon also has plans to offer in the future the ability to seamlessly move data from Glacier to S3 based on data service policies.
Data retrieval is priced differently. Amazon allows you to retrieve 5% of your storage for free per month. Retrieve more than 5% and you will be charged around $0.12 per Gigabyte. Retrieval requests take between 3-5 hours before the data is available for download.
Like most cloud storage offerings Amazon Glacier stores data using AES-256 encryption. If encryption is required on data in flight to the cloud, businesses can encrypt using their own keys prior to uploading.
Glacier can support various use cases, including the archiving of enterprise information, media assets, scientific data, database backups and data required to be held for regulatory compliance.
Most archival solutions for business require expensive capital expenditure, ongoing operational, support and maintenance costs, requires trained staff to operate and maintain and often leads to wasted capacity due to over provisioning for redundancy and unexpected growth.
Analysing the cost implications of using Glacier against a typical in house archiving solution, Glacier is a very attractive offering. The real value is that Amazon are offering to store your data essentially forever at a small cost. Amazon are taking all the heavy lifting away, no more need for complex capacity planning, dealing with multiple vendors for archival hardware and software upgrades and support, specialised training and offsite storage costs.
Amazon guarantees Glacier with an uptime of 99.9%, but also with a data reliability of 99.999999999%, meaning if you stored a billion objects you can expect to lose one every 100 years.
If your data archival needs must meet regulatory compliance you need to assess if Glacier datacentres meet those requirements. Glacier data can be stored in the US, Ireland (EU) or Japan.
Glacier will meet the needs of businesses with high volumes of “cold” data that must be stored for unspecified amounts of time where performance and restore speed are not important, but at the same time having the ability of simple retrieval for data that was stored 1, 5 or 10 years ago.