Big data tutorial: Everything you need to know
A comprehensive collection of articles, videos and more, hand-picked by our editors
Disk maker HGST has taken a bold step into the storage systems market with the launch of its Active Archive, a scale-out Amazon S3-compliant object storage system that marries helium-filled hard drives with storage software from Amplidata.
Active Archive starts with a 4.7PB rack and claims unlimited scalability. HGST is targeting the system at cloud provider, media, scientific, internet of things and big data use cases, where relatively slow data access times are acceptable and where its huge capacity – allied with efficient helium-filled drives – contribute to relatively low storage costs at scale.
The product comes in a 42U rack with up to 4.7PB of data, stored via Amplidata Himalaya object storage software with a maximum of 1.8 billion objects and a maximum object size of 16TB. Each 4.7PB rack offers 3.5GBps throughput and unlimited racks can be added in scale-out fashion. HGST bought Amplidata earlier this year.
Read more about object storage
- Object storage replaces tree-like file structures with flat systems of unique identifiers similar to DNS on the internet. In this ComputerWeekly.com guide we present step-by-step walkthroughs of the key elements of object storage.
- All but one of the big six storage vendors have object storage products that target public and private cloud environments and/or archiving use cases.
Helium for very large object storage
Helium is about 1/7th the density of air, which reduces friction against spinning components in the HDD when they start up. HGST said that, as the HDD runs, Helium brings a 36% decrease in power usage as well as the ability to run seven (thinner) platters in the drive rather than the usual five – an increase crucial to the improved capacity.
Object storage was developed to resolve challenges created by traditional hierarchical file systems when managing web-scale storage. In particular, that very large file systems with deep hierarchies become unwieldy when accessing one file from a set of billions.
Instead of storing data as files, object storage represents data as objects, each addressed by a unique identifier and stored in a flat address space with no sub-directories. Retrieval is via identifiers contained in an indexed database and assembled at a higher level into files.
Amplidata’s Himalaya object storage software is aimed at service providers and OEMs that need to access billions of objects with storage capacities of exabytes and beyond, in enterprise systems, and private and public clouds. Connectivity is via NFS/SMB via a NAS gateway to Amazon S3 calls. Amplistor uses erasure coding for resilience and claims 15 nines durability.
HGST targets OEMs with object storage with AWS OpsWorks